Handbook for Blast Resistant
Content
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
HANDBOOK FOR
BLAST-RESISTANT
DESIGN OF BUILDINGS
HANDBOOK FOR
BLAST-RESISTANT
DESIGN OF BUILDINGS
Edited by
Donald O. Dusenberry
JOHN WILEY & SONS, INC.
To my wife, Alice
∞
This book is printed on acid-free paper.
C 2010 by John Wiley & Sons, Inc. All rights reserved.
Copyright
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise,
except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA
01923, (978) 750-8400, fax (978) 646-8600, or on the web at www.copyright.com. Requests to the
Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons,
Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at
www.wiley.com/go/permissions.
Limit of Liability/Disclaimer of Warranty: While the publisher and the author have used their best
efforts in preparing this book, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this book and specifically disclaim any implied
warranties of merchantability or fitness for a particular purpose. No warranty may be created or
extended by sales representatives or written sales materials. The advice and strategies contained
herein may not be suitable for your situation. You should consult with a professional where
appropriate. Neither the publisher nor the author shall be liable for any loss of profit or any other
commercial damages, including but not limited to special, incidental, consequential, or other
damages.
For general information about our other products and services, please contact our Customer Care
Department within the United States at (800) 762-2974, outside the United States at (317)
572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print
may not be available in electronic books. For more information about Wiley products, visit our
web site at www.wiley.com.
Library of Congress Cataloging-in-Publication Data:
Handbook of blast resistant design of buildings / edited by Donald O. Dusenberry.
p. cm.
Includes index.
ISBN 978-0-470-17054-0 (cloth)
1. Building, Bombproof. I. Dusenberry, Donald O.
TH1097.H36 2010
693.8 54–dc22
2009019203
ISBN: 978-0-470-17054-0
Printed in the United States of America
10 9 8 7 6 5 4 3 2 1
CONTENTS
Preface
xv
Contributors
xix
I
DESIGN CONSIDERATIONS
1 General Considerations for Blast-Resistant Design
Donald O. Dusenberry
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
1.10
Introduction
Design Approaches
The Blast Environment
Structure As an Influence on Blast Loads
Structural Response
Nonstructural Elements
Effect of Mass
Systems Approach
Information Sensitivity
Summary
References
2 Design Considerations
Robert Ducibella and James Cunningham
2.1
2.2
2.3
2.4
Introduction
A New Paradigm for Designing Blast-Resistant Buildings,
Venues, and Sites
A Brief History of Recent Terrorist Attacks
2.3.1 Terrorists’ Use of Explosives
2.3.2 Vehicle-Borne Improvised Explosive Devices
2.3.3 Person-Borne Improvised Explosive Devices
2.3.4 Locally Available Explosives
2.3.5 Some Counterterrorism Considerations
Collaborating to Analyze Risk
2.4.1 Step 1—Threat Identification and Rating
2.4.2 Step 2—The Asset Value Assessment
2.4.3 Step 3—The Vulnerability Assessment
1
3
3
4
5
6
8
9
10
12
13
14
15
17
17
18
21
21
22
24
25
27
28
28
31
34
v
vi
3
CONTENTS
2.4.4 Step 4—The Risk Assessment
2.4.5 Step 5—Considering Mitigation Options
2.4.6 The Continuing Role of Risk Management
2.5 Consequence Management
2.5.1 Consequence Evaluation
2.5.2 Function Redundancy
2.5.3 Building Location
2.5.4 Building Dispersal/Distribution of Functional Programs
2.5.5 Disaster Recovery and Contingency Planning
2.6 Threat Reduction
2.6.1 Accidental Explosions
2.6.2 Intentional Explosions
2.7 Vulnerability Reduction
2.7.1 Standoff Distance
2.7.2 Physical Security
2.7.3 Operational Security
2.7.4 Structural Design
2.8 Risk Acceptance
2.8.1 Design to Threat
2.8.2 Design to Budget
2.9 Some Recent Examples of Security Design “Best Practices”
2.10 Related Phenomena
2.10.1 Progressive Collapse
2.10.2 Disruption of Evacuation, Rescue, and
Recovery Systems
2.10.3 Attendant Fires
2.11 Security Design Consideration Guidelines
2.12 Conclusion
References
38
39
40
42
44
48
51
54
56
57
59
60
63
64
65
65
65
70
71
73
75
76
77
Performance Criteria for Blast-Resistant Structural Components
Charles J. Oswald
87
3.1
3.2
3.3
3.4
87
88
91
3.5
Introduction
Building and Component Performance Criteria
Response Parameters
Empirical Correlations between Response Parameters
and Component Damage
Response Criteria Development
3.5.1 Explosive Safety Criteria
3.5.2 Response Criteria for Antiterrorism
3.5.3 Response Criteria for Blast-Resistant Design of
Petrochemical Facilities
3.5.4 Blast Resistant Doors
3.5.5 Blast-Resistant Windows
79
81
83
84
85
95
99
99
102
105
107
109
CONTENTS
3.5.6 Response Criteria for Equivalent Static Loads
3.5.7 Comparisons of Published Response Criteria
3.6 Response Criteria Limitations
References
4 Materials Performance
Andrew Whittaker and John Abruzzo
4.1
4.2
Introduction
Structural Steel
4.2.1 Stress-Strain Relationships
4.2.2 Constitutive Models for Structural Steel
4.2.3 Component Level Strain Rate and Temperature Effects
4.2.4 Mechanical Properties for Design
4.2.5 Failure Modes of Structural Components
4.3 Reinforced Concrete
4.3.1 Stress-Strain Relationships for Concrete
4.3.2 Stress-Strain Relationships for Reinforcement
4.3.3 Constitutive Modeling of Concrete and Rebar
4.3.4 Component Level Strain-Rate Effects
4.3.5 Mechanical Properties for Design
4.3.6 Component-Level Failure Modes
4.4 Strength-Reduction Factors for Steel and Reinforced Concrete
References
5 Performance Verification
Curt Betts
5.1
5.2
5.3
Introduction
Performance Verification
Testing
5.3.1 Vehicle Barrier Testing
5.3.2 Building Components
5.4 Analysis
5.5 Peer Review
References
vii
112
113
114
116
119
119
119
119
120
123
125
127
129
129
132
132
136
138
141
144
145
149
149
149
150
150
151
156
157
157
II BLAST PHENOMENA AND LOADINGS
159
6 Blast Phenomena
Paul F. Mlakar and Darrell Barker
161
6.1
6.2
6.3
Introduction
Sources of Blasts
Characteristics of Blast Waves
161
162
170
viii
7
8
CONTENTS
6.3.1 Key Parameters
6.3.2 Scaling
6.4 Prediction of Blast Parameters
6.4.1 High Explosives
6.4.2 Bursting Pressure Vessels
6.4.3 Vapor Cloud Explosions
6.5 Summary
References
170
171
172
172
177
178
181
181
Blast Loading
Paul F. Mlakar and William Bounds
183
7.1 Introduction
7.2 Empirical Method
7.2.1 Empirical Method—Basic Blast Wave Example
7.3 Front Wall Loads
7.3.1 Empirical Method—Front Wall Loading Example
7.3.2 Empirical Method—Oblique Angle Example
7.4 Side Wall and Roof Loads
7.4.1 Empirical Method—Side Wall Loading Example
7.4.2 Empirical Method—Roof Loading Example
7.5 Rear Wall Loads
7.5.1 Empirical Method—Rear Wall Loading Example
7.6 Confined Explosions
7.7 Leakage
7.8 Ray-Tracing Procedures
7.9 Summary
References
183
183
186
186
188
192
192
194
196
197
197
198
206
208
212
212
Fragmentation
Kim King
215
8.1 Introduction
8.2 Debris
8.3 Loadings
8.3.1 Primary Fragmentation
8.3.2 Secondary Fragmentation
8.4 Design Fragment Parameters
8.4.1 Fragment Final Velocity
8.4.2 Fragment Trajectory
8.5 Fragment Impact Damage
8.5.1 Fragment Penetration into Miscellaneous Materials
(THOR Equation)
8.5.2 Steel
8.5.3 Fragment Penetration into Concrete Targets
8.5.4 Fragment Perforation of Concrete Targets
215
215
215
216
218
226
226
227
228
229
231
233
235
CONTENTS
8.5.5 Fragment Spalling of Concrete Targets
8.5.6 Roofing Materials
8.5.7 Other Materials
References
III SYSTEM ANALYSIS AND DESIGN
9
236
236
237
237
239
Structural Systems Design
Robert Smilowitz and Darren Tennant
241
9.1
241
241
243
244
245
246
248
251
252
252
253
255
255
256
258
258
261
9.2
9.3
9.4
10
ix
General Discussion
9.1.1 Seismic versus Blast
9.1.2 Analytical Methods
Modeling
9.2.1 Systems
9.2.2 Materials
9.2.3 Members
9.2.4 Connections
Analytical Approaches
9.3.1 P-I Diagrams
9.3.2 Single-Element Analyses
9.3.3 Structural Systems Response
9.3.4 Explicit Dynamic Finite Element Analyses
Progressive Collapse
9.4.1 European Guidance
9.4.2 U.S. Guidance
References
Building Envelope and Glazing
Eve Hinman and Christopher Arnold
263
10.1 Design Intent
10.1.1 Life Safety
10.1.2 Emergency Egress and Facilitating Search
and Rescue
10.1.3 Critical Functions (Protecting Equipment and
Business Processes)
10.2 Design Approach
10.2.1 Response Criteria
10.2.2 Static versus Dynamic
10.2.3 Balanced Design
10.2.4 Load Path
10.3 Fenestration
10.3.1 Glass
10.3.2 Mullions/Transoms
10.3.3 Frame and Anchorage
263
263
264
264
265
269
270
270
270
272
273
278
279
x
CONTENTS
10.4
10.5
10.6
10.7
11
12
10.3.4 Supporting Structure
10.3.5 Other Penetrations
Exterior Walls
10.4.1 Concrete Walls
10.4.2 Masonry
10.4.3 Steel
10.4.4 Other
Roof Systems
10.5.1 Concrete
10.5.2 Steel
10.5.3 Composite
10.5.4 Penthouses/Gardens
Below Grade
Reduction of Blast Pressures
References
280
280
281
282
285
285
286
289
289
289
290
290
290
292
294
Protection of Spaces
MeeLing Moy and Andrew Hart
297
11.1 Areas Isolating Interior Threats
11.2 Stairwell Enclosures
11.3 Hardened Plenums
11.4 Safe Havens
11.4.1 FEMA Documents
11.4.2 Multi-Hazard Threats
11.4.3 Design Requirements for Protective
Shelters
References
297
298
298
299
299
300
Defended Perimeter
Joseph L. Smith and Charles C. Ellison
307
12.1 Goals
12.2 Standoff
12.2.1 Balancing Hardening with Standoff
12.2.2 Balancing Costs
12.2.3 Site Planning
12.3 Vehicle Control Barriers
12.3.1 Crash Testing
12.3.2 Crash Modeling
12.3.3 Walls
12.3.4 Bollards
12.3.5 Active Wedge
12.3.6 Beam Barriers
12.3.7 Cable-Based Systems
307
307
309
311
313
316
316
317
319
319
320
320
323
301
305
CONTENTS
12.3.8
12.3.9
13
xi
Planter and Surface Barriers
Berms, Ditches, and Other Landscaping
Features
12.4 Pedestrian Control Barriers
12.5 Blast Walls and Berms
References
324
Blast-Resistant Design of Building Systems
Scott Campbell and James Ruggieri
331
13.1 Background
13.2 Introduction
13.3 Design Considerations
13.3.1 Level of Protection
13.3.2 Blast Pressures
13.3.3 Shock Induced by the Structure
13.3.4 Equipment/System Anchorage
13.3.5 Placement of Critical Systems Equipment and
Control Stations
13.3.6 Staffing and Building Operations
13.3.7 Construction of Hardened Spaces
13.3.8 HVAC and Plumbing Systems
13.3.9 Electrical Systems
13.3.10 Lighting Systems
13.3.11 Other Systems/Considerations
13.4 Loading Calculation
13.4.1 Blast Pressure
13.4.2 In-Structure Shock
13.5 Summary
References
331
332
333
334
334
335
337
324
325
327
329
340
340
341
341
344
346
346
348
349
352
362
363
IV BLAST-RESISTANT DETAILING
365
14
Blast-Resistant Design Concepts and Member Detailing
Steven Smith and W. Gene Corley
365
14.1 General
14.1.1 Scope
14.2 Failure Modes
14.2.1 Flexural
14.2.2 Diagonal Tension
14.2.3 Direct Shear
14.2.4 Membrane
14.2.5 Stability
367
367
368
368
369
369
369
370
xii
15
CONTENTS
14.3 Detailing
14.3.1
General
14.3.2
Splices
14.3.3
Columns
14.3.4
Beams
14.3.5 Beam-Column Joints
14.3.6
Slabs
14.3.7
Walls
References
370
370
371
372
375
377
378
380
380
Blast-Resistant Design Concepts and Member Detailing: Steel
Charles Carter
383
15.1 General
15.1.1 Typical Building Designs
15.1.2 Prescriptive Building Designs
15.1.3 Performance-Based Building Designs
15.2 Blast Effects on Structural Steel and Composite Structures
15.2.1 Member Ductility
15.2.2 Connection Ductility
15.2.3
Overstrength
15.2.4 Beneficial Strain-Rate Effects
15.2.5 Beneficial Effects of Composite Construction
15.2.6 Perimeter Column Design
15.2.7 Perimeter Girder Design
15.2.8
Slab Design
15.3 Analysis and Design of Structural Members
15.4 Steel Material Properties for Blast Design
15.4.1 Strength Increase Factor (SIF)
15.4.2 Dynamic Increase Factor (DIF)
15.4.3 Dynamic Design Stress
15.5 Design Criteria for Blast Design
15.5.1
General
15.5.2 Load Combinations
15.5.3 Resistance Factor and Factor of Safety
15.5.4 Local Buckling
15.5.5 Lateral-Torsional Buckling
15.5.6 Deformation Criteria
15.5.7 Detailing for Specific Failure Modes:
15.6 Examples
15.6.1 Example 1—Determining Capacities
15.6.2 Example 2—Design and Analysis for Blast Loads on
Members
15.7 Design of Connections
References
383
383
384
385
386
386
386
386
386
387
387
387
388
388
388
389
389
390
390
390
391
391
391
391
391
393
397
397
402
418
419
CONTENTS
16
17
xiii
Blast-Resistant Design Concepts and Member
Detailing: Masonry
Shalva Marjanishvili
421
16.1 General Considerations
16.1.1 Masonry
16.1.2 Reinforcement
16.1.3 Mortar
16.1.4 Grout
16.1.5 Construction Methods
16.2 Failure Modes
16.2.1 Flexure
16.2.2 Diagonal Tension Shear
16.2.3 Direct Shear
16.2.4 Breach and Spall Phenomena
16.3 Reinforced Masonry Detailing
16.3.1 General
16.3.2 Longitudinal Reinforcement
16.3.3 Horizontal Reinforcement
16.3.4 Walls
16.3.5 Support Connections
16.4 Unreinforced Masonry
16.4.1 Performance Evaluation
16.4.2 Retrofit Recommendations
References
423
424
424
425
425
425
426
428
431
432
432
434
435
435
435
438
438
439
439
440
442
Retrofit of Structural Components and Systems
John E. Crawford and L. Javier Malvar
445
17.1 Introduction
17.2 Retrofit of Columns
17.2.1 Reinforced Concrete Columns
17.2.2 Steel Columns
17.3 Retrofit of Walls
17.3.1 Masonry Walls
17.3.2 Stud Walls
17.4 Floors
17.5 Beams/Girders/Connections
17.6 Structural System
17.7 References
17.7.1 Inexact Science
17.7.2 Complexities
References
445
446
446
454
458
458
466
466
468
469
469
469
470
470
Index
477
PREFACE
The need for protection against the effects of explosions is not new. The use
of explosive weaponry by the military necessitated resistive entrenchments ages
ago. Industrialization of our societies well over a century ago meant that we
intended to manufacture, store, handle, and use explosives in constructive ways.
To support these military and industrial purposes, a relatively small group of
designers have worked to devise ways to strengthen the blast resistance of our
structures.
Early attempts at blast-resistance design necessarily relied on judgment, test,
and trial-and-error construction to find the best solutions. As technology improved, designers became better able to predict the influences of explosions
and the resistive responses that they strove to impart into their designs. More
recently, in the past several decades chemists, physicists, blast consultants, and
structural engineers have been empowered by technologies and computational
tools that have enhanced the precision of their analyses and the efficiency of their
designs.
At the same time, the need has increased. The small contingent of designers
skilled in the art and science of creating structural designs that will resist explosive forces has been joined by a larger group of architects, engineers, blast
consultants, and security consultants who are trying to respond to the increasing concern from a broader group of clients who fear an exposure that they did
not anticipate before and frequently did not bring upon themselves. Consultants
who have never before had to assess risks, devise risk-reduction programs, provide security systems, establish design-base threats, calculate the pressures and
impulses from explosions, and create cost-effective structural designs are being
thrust into the process. Many are ill-trained to respond.
There are several good references on some of the aspects of designing for
blast resistance. Some of these references support military purposes or for other
reasons have government-imposed restrictions against dissemination. As such,
they are not widely available to consultants working in the private sector. Nearly
all those references and the references that are public each treat an aspect of
blast phenomenology, security systems, and structural design for blast resistance,
but few, if any, bring together in one place discussions of the breadth of the
issues that are important for competent designs. Consultants are forced to collect
a library of references and extract from each the salient information that they
then synthesize into a comprehensive design approach.
xv
xvi
PREFACE
In addition, practitioners who do receive the limited-distribution references
for the first time or who find references that are public usually discover
immediately that designing for blast resistance is completely different from designing for any environmental load they encountered previously. Designers often
realize quickly that they are embarking on design process for which they do not
have the knowledge or experience for adequate competency. Those who do not
have this realization might be operating at risk if they are not careful and thorough students.
The purpose for this handbook is to bring together into one publication discussions of the broad range of issues that designers need to understand if they
are to provide competent, functional, and cost-efficient designs. The contributors
to this book are among the most knowledgeable and experienced consultants and
researchers in blast resistant design, and contribute their knowledge in a collaborative effort to create a comprehensive reference. Many of the contributors to
this handbook are collaborating in the development of the first-ever public-sector
standard for blast resistant design, being developed contemporaneously with this
handbook by the Structural Engineering Institute (SEI) of the American Society of Engineers. While there undoubtedly will be some differences between the
SEI standard and this handbook, many readers will consider these publications
as companions.
This handbook is organized into four parts, each addressing a range of aspects
of blast-resistance design.
Part 1: Design Considerations provides an overview of basic principles.
It has five chapters dealing with general considerations and the design process; risk analyses, reduction, and avoidance; criteria that establish acceptable performance; the science of materials performance under the extraordinary
blast environment; and performance verification for technologies and solution
methodologies.
Part 2: Blast Phenomena and Loadings, in three chapters, describes the
explosion environment, loading functions to be used for blast response analysis,
and fragmentation and associated methods for effects analyses.
Part 3: System Analysis and Design has five chapters that cover analysis and design considerations for structures. This part instructs on structural, building envelope, component space, site perimeter, and building system
designs.
Part 4: Blast-Resistant Detailing addresses detailing structural elements for
resistance. Chapters on concrete, steel, and masonry present guidance that is
generally applicable for new design. The fourth chapter addresses retrofits of
existing structures.
I wish to thank all the contributors for their commitment to this work, their
collaborative spirit, and, of course, their willingness to share the blast-related
expertise that they have presented in their chapters. I wish to thank Steven Smith
of CTLGroup in particular, for organizing and harmonizing the four chapters of
Part 4. William Zehrt of the Department of Defense Explosives Safety Board
improved the quality of this handbook by reviewing the chapters of Part 2.
PREFACE
xvii
I also wish to thank James Harper, Editor of John Wiley & Sons for supporting this effort; Daniel Magers, Senior Editorial Assistant, and Amy Odum
for her able supervision of the copyediting and production; and the copyeditors,
compositors, typesetters, and others of the publisher’s staff who have professionally assembled this book and brought it to publication.
Donald O. Dusenberry
Wakefield, Massachusetts
CONTRIBUTORS
John Abruzzo, P.E.
Principal
Thornton Tomasetti
555 12th Street, Suite 600
Oakland, CA 94607
Tel: (510) 433-9370
JAbruzzo@ThorntonTomasetti
.com
Christopher Arnold, FAIA, RIBA
President, Building Systems
Development Inc
Palo Alto CA
1248 Waverley Street
Palo Alto, CA 94301
Tel: 650-462-1812
[email protected]
[email protected]
Curt P. Betts, P.E.
Chief, Security Engineering Section
US Army Corps of Engineers
Protective Design Center
1616 Capitol Avenue
Omaha, Nebraska 68102-4901
Tel: (402) 995-2359
[email protected]
Darrell D. Barker, P.E.
Vice President
Extreme Loads and Structural
Risk
ABS Consulting
14607 San Pedro Ave., Suite 215
San Antonio, Texas 78232
Tel: (210) 495-5195
[email protected]
William Bounds, P.E.
Fluor
PO Box 5014
Sugar Land, Texas 77487
[email protected]
Scott Campbell, Ph.D., P.E.
Structural Analysis Consulting
Group
PO Box 91364
Louisville, KY 40291
Tel: (502) 762-9596
[email protected]
Charles Carter
American Institute of Steel
Construction
One East Wacker Drive
Suite 700
Chicago, Illinois 60601-1802
Tel: (312) 670-2400
[email protected]
W. Gene Corley, Ph. D.
CTLGroup
5400 Old Orchard Road
Skokie, Illinois 60077-4321
Tel: (847) 972-3060
Fax: (847) 965-6541
[email protected]
xix
xx
CONTRIBUTORS
John E. Crawford
Karagozian & Case
2550 N. Hollywood Way, Suite 500
Burbank, CA 91505-5026
Tel: (818) 240-1919
[email protected]
James D. Cunningham
Williamsburg, Virginia
Tel: (757) 645-4057
[email protected]
Robert Ducibella
Senior Principal
Ducibella Venter & Santore
Security Consulting Engineers
Sturbridge Commons – Franklin
House
250 State Street
North Haven, CT 06473
Tel: (203) 288-6490
[email protected]
Donald O. Dusenberry, P.E.
Senior Principal
Simpson Gumpertz & Heger Inc.
41 Seyon Street, Building 1, Suite 500
Waltham, MA 02453
Tel: (781) 907-9000
[email protected]
Chuck Ellison, P.E.
Senior Security Engineer
Applied Research Associates, Inc.
119 Monument Place
Vicksburg, MS 39180
Phone: (601)638-5401
[email protected]
Andrew Hart, Ph.D., MSc, BEng
(Hons), Aff.M.ASCE
Martinez, CA 94553
Tel: (925) 370-3866
[email protected]
Eve Hinman, Eng.Sc.D., P.E.
President
Hinman Consulting Engineers, Inc.
One Bush Street, Suite 510
San Francisco, CA 94104
Tel: (415) 621-4423
[email protected]
Kim W. King, P.E.
Director of Engineering
2195 Redwoods Crest
San Antonio, TX 78232
Tel: (210) 213-3737
[email protected]
L. Javier Malvar
Naval Facilities Engineering Service
Center
1100 23rd Avenue
Port Hueneme, CA 93043-4370
[email protected]
Shalva M. Marjanishvili, Ph.D.,
P.E., S.E.
Technical Director
Hinman Consulting Engineers, Inc.
One Bush Street, Suite 510
San Francisco, CA 94104
Tel: (415) 621-4423
[email protected]
Paul F. Mlakar, Ph.D., P.E.
U.S. Army Engineer Research and
Development Center
3909 Halls Ferry Road
Vicksburg, MS 39180
Tel: (601) 634-3251
[email protected]
MeeLing Moy, P.E.
President
The Link CE, PLLC
New York
Tel: (646) 385-5096
[email protected]
CONTRIBUTORS
Charles J. Oswald, Ph.D, P.E.
Protection Engineering Consultants
4203 Gardendale, Suite C112
San Antonio, TX 78229
Tel: (512) 380-1988
[email protected]
James Angelo Ruggieri, P.E.
10710 Timberidge Road
Fairfax Station, VA 22039
Tel: (703) 250-3671
[email protected]
Robert Smilowitz, Ph.D., P.E.
Weidlinger Associates, Inc.
375 Hudson Street
12th Floor
New York, New York 10014-3656
Tel: (212) 367-3090
[email protected]
Joseph L. Smith, PSP
Senior Vice President
Applied Research Associates, Inc.
119 Monument Place
Vicksburg, MS 39180
Tel: (601) 638-5401
[email protected]
Steven Smith
CTLGroup
10946 Eight Bells Lane
Columbia, MD 21044
Tel: (410) 997-0400
[email protected]
Darren Tennant
Weidlinger Associates, Inc.
6301 Indian School Road, NE,
Suite #501
Albuquerque, NM 87110
Tel: (505) 349-2820
[email protected]
Andrew Whittaker, Ph.D., S.E.
University at Buffalo
Buffalo, NY 14260
Tel: (716) 645-4364
[email protected]
xxi
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
I
Design Considerations
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
1
General Considerations for
Blast-Resistant Design
Donald O. Dusenberry
1.1 INTRODUCTION
Until recently, relatively few engineers and architects have had to design structures and their systems to resist the effects of explosions. Military engineering
personnel, consultants to the federal government, and consultants to industries
that use explosive or volatile materials constituted the primary population of designers routinely analyzing blast effects.
Following the explosion that demolished the Alfred P. Murrah Federal Building in Oklahoma City in 1995, members of the structural design and construction
industries have been increasingly quizzed by owners about blast-related hazards,
risks, and methods of protection. The types and numbers of clients seeking blast
resistance in their structures have expanded.
The terrorist events of the recent past and the fear that others may occur in the
future have led many businesses, particularly those with an international presence, to consider their vulnerability. And, of course, as their neighbors work to
enhance the performance of their buildings, owners and tenants who do not envision themselves as targets of malevolent acts nevertheless begin to wonder if
their structures might be damaged as a consequence of their proximity to targets. Some have argued that adding blast resistance and enforcing standoff for
one building on a block unfortunately increases the threat for others, because it
encourages aggressors to attempt to assemble bigger bombs and detonate them
closer to the target’s neighbors.
There seems to be a sense of anxiety about the vulnerability of our buildings,
bridges, tunnels, and utilities in the midst of numerous recognized international
social and political instabilities, and given the potential for domestic groups and
individuals to seek influence and create disruption by resorting to violent means.
As a result, consultants designing rather pedestrian buildings now are expected to
provide advice and sometimes specific enhancements in response to quantifiable
threats, as well as perceived vulnerabilities.
In this environment, engineers need training and information so that they can
provide designs that effectively enhance a building’s response to explosions.
3
4
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.2 DESIGN APPROACHES
Most engineers and architects serving clients with growing interest in blast resistance are uninitiated in the relevant design practice. Blast loading is very different from loadings commonly analyzed by structural engineers. Peak pressures
are orders of magnitude higher than those associated with environmental loads,
but their durations generally are extremely short compared to natural periods of
structures and structural components. In addition, given that the risk of an explosion at any one facility normally is very low and the costs to achieve elastic
response often are prohibitive, designs usually engage the energy-dissipating capability of structural and enclosure elements as they are deformed far into their
inelastic ranges. This forces engineers to account for geometric and material
nonlinearities.
At first, designing for blast resistance might sound similar to designing for
seismic resistance because neither is static and both rely on post-yield response.
But even those similarities are limited. The dominant frequencies of seismic excitations are on the order of the lowest natural frequencies of building response,
not much faster, as is generally the case for blast loadings. Blast loading usually
is impulsive, not simply dynamic.
While we tolerate some damage in earthquakes, to dissipate energy, we usually allow more damage for blast events. We expect facades to sustain severe
damage. In fact, blast-resistant design often tolerates breaching of the building
enclosure (with attendant risk of fatalities) and even sometimes partial collapse
of buildings.
Many blast-resistant designs require very sophisticated approaches for the
analysis of building response to explosions (National Research Council 1995).
There are techniques for accurate assessment of blast pressures and impulses
in complicated environments, modeling the influence of those blast loadings on
surfaces, and structural response to those loads. There are critical facilities and
blast conditions that warrant the use of these techniques. However, much blastresistant design is performed following simplified procedures (U.S. Department
of Defense 2008) that approximate actual conditions, and therefore lack high
fidelity. This often is appropriate because of, and at least in part follows from,
inevitable uncertainties that mask the phenomenon and the structure’s response.
In addition, there are practical matters of prudence, economics, and risk acceptance that drive analyses of blast response.
Risk analyses are important components of the design for blast resistance
(Federal Emergency Management Agency 2003). Among the products of such
analyses are estimates of the threat for which a structure should be designed. The
magnitude of intentional, nonmalevolent explosions and industrial explosions
sometimes can be estimated with precision commensurate with that of other
common loadings (Center for Chemical Process Safety 1996). The quantity of
explosive materials can be estimated, the potential locations of the design-base
explosion can be isolated, and often there are relatively few nearby objects that
significantly affect the shock front advance.
THE BLAST ENVIRONMENT
5
This is not the case for many accidental explosions and most malevolent explosions. The assessment of the threat in these instances often does not have a
probabilistic base. When sufficient data do not exist, consultants are forced to
use judgment rather than hard science to establish the threat.
When data are not available, consultants often establish the magnitude of the
threat of a malevolent explosion by assessing the probable size of the container
(e.g., letter, satchel, package) in which a bomb is likely to be delivered (U.S.
Department of Defense 2002a), and then selecting a design-base explosive mass
based on a fairly arbitrary assignment of the quantity of explosive that could
reasonably be accommodated in that container. In these cases, there is relatively
high uncertainty about the intensity of the explosion that might actually occur.
Obviously under these circumstances, there is a commensurate level of uncertainty about the outcome.
1.3 THE BLAST ENVIRONMENT
Engineers skilled in the design of buildings for occupancy-related and environmental loads (e.g., dead, live, wind, snow, and seismic loads), but faced with a
new challenge to design for blast loads, often find themselves ill-equipped for
the challenge. Designers are used to treating all other common loadings as either
static or quasi-static, because the rise time and duration for the equivalent load
are on the order of, or longer than, the longest natural periods of the structure.
Designing for blast loading generally cannot follow this approach.
Conventional design for common time-varying loads, including wind and
seismic, includes techniques that allow conversion of these dynamic phenomena
into quasi-static events that recognize and simplify the dynamics. Wind loads
defined in one of the most common references (American Society of Civil Engineers 2005) are based on an acknowledgment of the range of natural frequencies
of common structural frames, and are calibrated to those values. When the frequency of a subject building falls outside of that default range, common design
approaches provide for specified adjustments to the quasi-static design loads to
account for dynamic response.
Common seismic design (American Society of Civil Engineers 2005) involves
a very elaborate conversion of the dynamic loading environment into a quasistatic analysis problem. Building systems are characterized for stiffness and ductility, and site conditions are evaluated for seismic exposures and characteristics
of shaking. On the basis of extensive research into building performance and a
fair amount of cumulative experience evaluating the actual earthquake response
of designed structures, the complicated loadings—which are as much a function
of the building design as they are of the environment in which structures are
built—are idealized as a series of externally applied loads that are thought to
mimic the loading effects of an earthquake. Complicated though the approach is,
many buildings can be designed for earthquakes by engineers with little familiarity with dynamic behavior.
6
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
Our conventional approach to blast design is similar to that for seismic design,
in two important ways: (1) both loadings clearly are dynamic and, hence, solutions are energy-based, and (2) the way we detail structural elements determines
the effective loads for which structures must be designed (meaning, we limit the
strength we need to supply by allowing post-elastic behavior to dissipate energy).
However, blast loading, with its extremely fast rise time and usually short duration, is either dynamic or impulsive, depending on the nature of the explosive, its
distance from the subject structure, and the level of confinement that the structure
creates for the expanding hot gases (Mays and Smith 1995).
The impulsive form of the very fast load rise time and very short load duration normally associated with blast loading requires analytical approaches that
generally demand direct solution of energy balance equations (U.S. Department
of Defense 2008; Mays and Smith 1995).
1.4 STRUCTURE AS AN INFLUENCE ON BLAST LOADS
The pressure and duration of the impulse associated with a blast are influenced
by reflections of the shock front (U.S. Department of Defense 2008). Reflection
sources include the ground below the detonation point and building surfaces that
have sufficient mass or ductility to remain largely in place for the duration of the
impulse. When shock fronts are reflected, their pressures are magnified as a function of the proximity, robustness, and material characteristics of the impacted object (Bangash 1993). The more robust that object, the greater the reflected energy
because less energy is dissipated by the response (such as ground cratering) of
the surface. These variations often are neglected in conventional design.
For instance, facades normally are designed on the assumption that they are
perfect reflectors of the shock front. Designers following common procedures
are assuming that the facade components remain stationary for the duration of the
impinging shock front, causing peak pressures and impulses sufficient to reverse
the direction of the shock front. In practice, there can be some displacement of
the facade during the loading cycle. This displacement reduces the effectiveness
of the reflector, and correspondingly the impulse.
Analyses for interior explosions have additional complications, as designers
attempt to deal with the multiple reflections of the shock front within the structure, and pressures that develop from containment of expanding hot gases (Mays
and Smith 1995)–a phenomenon normally neglected for external explosions.
Further, the geometry of the confining volume and the location of the explosion within the volume can substantially affect the pressures on surfaces (U.S.
Department of Defense 2008). The science that describes the pressure history on
interior surfaces is complex, and not generally considered rigorously in common
blast-resistant design processes.
Providing blast venting through frangible components to mitigate the effects
of interior explosions is even more complex, since the release time for the venting
component is a key, but difficult to assess, factor in the determination of the
STRUCTURE AS AN INFLUENCE ON BLAST LOADS
7
magnitude of the pressure buildup. Approximations usually govern the analyses
(U.S. Department of Defense 2008).
Clearly, there is interplay between the performances of building facades and
frames. While in most cases the primary reason we enhance the performance
of a facade is to protect occupants, we gain protection for the structure as well.
Blast shock fronts that are not repelled by the facade will advance into a building,
inducing pressures on interior surfaces of the structure and threatening interior
columns, walls, and floor systems. Blast-related upward impulses on floor slabs
can reverse force distributions in these structural elements. In systems that are
not strong and ductile enough for these reversed forces, blast-induced deflections
can fracture structural elements that are required to resist gravity loads. Hence,
floor systems can fail after the direct effects of the blast pass and the slab falls
back downward under the influence of gravity.
Of course, by designing the facade to resist the effects of an explosion, the designer is forcing the structure to become a support for the blast loads. Depending
on the performance criteria, designers need to demonstrate that the framing system can support the applied loads, and that the structure as a whole will remain
standing with an acceptable level of damage.
Building enclosures normally are designed to resist blast effects by inelastic
flexural action, but it is possible to design facades to resist blast effects through
catenary action as well. In particular, blast retrofits sometimes include new
“catch systems” that are intended to reduce intrusion of blast pressures and creation of lethal missiles, by acting as a net inside the original exterior wall system.
In any case, the lateral displacement of the system often is large enough to
open gaps between wall panels or between panels and floor slabs. When this
happens, there is potential for leakage pressures to enter the building (U.S. Department of Defense 2002b), even when windows stay in place. This is particularly true in response to large, relatively distant explosions that have relatively
long-duration impulses.
Pressure fronts that leak past facades that are damaged but remain in place
normally are assumed to have insufficient energy to induce significant damage to
interior structural components. However, these leakage pressures can cause personal injuries and damage to architectural and mechanical systems if they are not
designed for resistance.
Add to the effects of leakage pressures the possibility that structural and architectural features on the inward-facing surfaces of facade components can
become missiles when the facade sustains damage as it deforms, and there remains substantial risk to occupants inside blast-resistant buildings even with
well-developed designs.
It is well established that breached fenestration leads to lethal missiles and
internal pressurization (American Society of Civil Engineers 1999). Common
design for blast resistance for malevolent attacks often is based on the premise
that a significant fraction of the fenestration in a building will fail (General Services Administration 2003). This is due in part to the variability of the properties
of glass, but also results from risk acceptance that employs the philosophy that
8
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
an explosion is unlikely and that full, “guaranteed” protection is prohibitively
difficult or expensive.
Hence, the effects of leakage pressures and missiles that are the product of
building materials fracturing in response to a blast often can be destructive to the
interiors of buildings, even when the facades of those buildings are designed to
resist the effects of an explosion. Except when the most restrictive approaches to
blast-resistant design are employed (e.g., with elastic response, so a building can
remain functional), parties with standing in the design process need to understand that substantial interior damage and occupant injuries are possible should
the design-base explosion occur.
1.5 STRUCTURAL RESPONSE
The shock front radiating from a detonation strikes a building component, it is
instantaneously reflected. This impact with a structure imparts momentum to exterior components of the building. The associated kinetic energy of the moving
components must be absorbed or dissipated in order for them to survive. Generally, this is achieved by converting the kinetic energy of the moving facade component to strain energy in resisting elements. Following the philosophy that blast
events are unusual loading cases that can be allowed to impart potentially unrepairable damage to structures, efficiency in design is achieved through post-yield
deformation of the resisting components, during which energy can be dissipated
through inelastic strain.
Of course, this means that the components that need evaluation often are deformed far beyond limits normally established for other loading types, and many
of the assumptions that form the basis for conventional design approaches might
not be valid. For instance, recognition of the extreme damage state normally
associated with dissipation of blast energy has led to debate about appropriate
values of the strength reduction factors ( factors) to be used for design.
In conventional design (American Concrete Institute 2005; American Institute of Steel Construction 2005), the nominal strengths of structural elements
are reduced by factors to account for uncertainty in the actual strength of
the elements, and for the consequences of failure. Their magnitudes for conventional design have been developed based on studies of structural responses
that are commensurate with service performance of buildings and, for seismic design, responses that are anticipated to be sufficiently limited and ductile to allow elements to retain most of their original load-carrying capacity.
Blast resistance, on the other hand, often takes structural elements far into
the inelastic range, to where residual strengths might be reduced from their
peaks, and alternative load-carrying mechanisms (e.g., catenary action) are engaged. Sometimes, designers anticipate complete failure of certain elements
if they are subjected to the design-base event. In this environment, it is not
at all clear that factors developed for conventional, nonblast design are
relevant.
NONSTRUCTURAL ELEMENTS
9
Common blast-resistant design often takes the values of the factors to be 1.0
(U.S. Department of Defense 2008). The bases for this approach range from the
uncertainty about what the actual values ought to be to the observation that loads
we assume for blast-resistant design are sufficiently uncertain that precision in
the values for is unjustified. It is further prudent to assume =1.0 when
performing “balanced design,” in which each structural element in a load path
is designed to resist the reactions associated with the preceding element loaded
to its full strength. Using =1.0 for determination of the full strength of the
elements in the load path tends to add conservatism to the loads required for the
design of the subsequent elements.
On the other side of the equation, designers often apply load factors equal to
1.0 to the blast effects (U.S. Department of Defense 2008). This follows from the
lack of a probabilistic base from which to determine the design threat, and the rationale that conservatism can be achieved by directly increasing the design threat.
In any event, the absence of complete agreement on how to address strength
reduction factors, and the valid observation that blast threats—particularly for
malevolent explosions—generally are difficult to quantify, reduce our confidence
in our ability to predict structural response with precision.
It is common in blast-resistant design to treat individual elements as singledegree-of-freedom nonlinear systems (U.S. Department of Defense 2008). Performance is judged by comparison of response to limiting ductility factors (i.e.,
the ratio of peak displacement to displacement at yield) or support rotations,
with the response calculated as though the structural element were subjected to
a pressure function while isolated from other structural influences. Of course,
much more sophisticated approaches are pursued for critical structures and complicated structural systems. However, research on structural response for very
high strain rates and very large deformations is limited, and results often are not
widely disseminated. In many respects, the sophisticated software now available
makes it possible to analyze with precision that exceeds our understanding of
structural response.
Hence, the simplified, single-degree-of-freedom approach forms the basis for
many designs. This approach usually is consistent with the precision with which
we model the blast environment and our knowledge of element behavior, but it
generally identifies the true level of damage only approximately. When considering elements as components of structural systems under the influence of blast, the
response of the individual elements can differ significantly from that determined
by analyses in isolation.
1.6 NONSTRUCTURAL ELEMENTS
Designers usually assume that the blast resistance of a structure is derived from
the elements that they design for this purpose. While this clearly is true in
large measure, in actual explosions, nonstructural elements—components disregarded in blast design—can act to reduce damage in a structure. It usually is
10
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
conservative, and therefore prudent, to ignore these components because the designer cannot be certain about the reliability, or even the long-term existence, of
building components that are not part of the structural design.
Nevertheless, elements with mass and ductility that stand between an explosion site and a target area can act to dissipate energy as they fail from the effects
of the blast. In fact, designers sometimes do rely on specific sacrificial elements
to reduce the blast effects on critical structural elements. The bases for this consideration are twofold: (1) through its failure, the sacrificial element dissipates
energy that would otherwise be imparted to the structural element, and (2) for the
brief time that the sacrificial element stays in place, it acts to reflect the shock
front, thereby reducing the impulse felt by the protected structural element. For
near-range conditions, when a bomb might otherwise be placed essentially in
contact with a key structural element, a sacrificial element such as an architectural column enclosure can enhance survivability simply by inhibiting close
placement of the explosive.
Of course, any shielding element that has inadequate strength, ductility, and
connection to remain attached to resisting elements is likely to become a missile. Some of the energy these elements absorb is dissipated through strain, but
the rest is retained as kinetic energy. The hazards created by these flying elements end only when that kinetic energy is brought to zero. Furthermore, care
is needed in the evaluation of the value of shielding elements that are not positioned closely to the structure under consideration, since shock fronts reform
beyond such elements, mitigating the protective value of the shield.
1.7 EFFECT OF MASS
The first influence of gravity comes to play when assessing the weights that
the designer assumes are present in the structure at the time of an explosion.
These weights, which are derived from the structure itself and its contents, act
concurrently with the explosion-induced loadings. As a result, they “consume”
some of the resisting capacity of the elements that are designed to resist the
explosion. In addition, for the most part, they remain on the structure after the
explosion and therefore must be supported by the damaged structure. The postblast distribution of these weights often will be uncertain.
On the beneficial side, mass often augments the blast resistance of structural
elements. Blast effects usually are impulsive, meaning that they impart velocity
to objects through development of momentum. With momentum being proportional to the product of mass and velocity (Eq. 1-1), and kinetic energy being
proportional to the product of mass and velocity squared, the larger the mass, the
smaller the velocity and, hence, the smaller the energy that must be dissipated
through strain (Eq. 1-2).
t1
(1.1)
I = F (t) dt = M V
t0
EFFECT OF MASS
where:
11
I = impulse
F(t) = time-varying force
t = time
M = mass
V = velocity
Ek =
1
1 I2
MV 2 =
2
2M
(1.2)
where: Ek = kinetic energy
Gravity also must be considered when elements or overall structures deform.
Vertical load-carrying elements often are designed to resist simultaneous vertical and lateral loads. Even when columns are not part of a structure’s lateral
load resisting system, it is common for them to be designed for an eccentricity
of the vertical load to account for inevitable moments that will develop in use.
Sometimes the magnitude of the eccentricity causing moment is assumed to be
on the order of 3% to 10% of the element’s cross section dimension (American
Concrete Institute 2005). Response to blast often deforms vertical structural elements far more than limits assumed for conventional design. The designer needs
to evaluate the ability to resist the resulting P- effects, both for individual elements and for the structure overall.
Structures as a whole generally are not pushed over by a common explosion.
The overall mass of a structure usually is large enough to keep the kinetic energy
imparted to the structure as a whole small enough that it can be absorbed by the
multiple elements that would need to fail before the building topples.
In many explosions that cause extensive destruction, the damage develops
in two phases: (1) the energy released by the explosion degrades or destroys
important structural elements, and (2) the damaged structure is unable to resist
gravity and collapses beyond the area of initial damage. In some of the most
devastating explosions, most of the structural damage has been caused by gravity
(Federal Emergency Management Agency 1996, Hinman and Hammond 1997).
Normally, individual elements fail, necessitating the activation of alternative
load paths within the structure to carry the gravity loads that remain after the
direct effects of the blast pass. Studies that assess these alternate load paths need
to consider the dynamic application of the redirected internal forces, as the sudden removal of load-carrying elements implies a change in potential energy, as
portions of a structure begin to drop. This change in potential energy necessarily imparts kinetic energy that must be converted to strain energy for the falling
mass to be brought to rest.
Hence, the evaluation of the full effect of a blast does not end with calculations
of blast damage to individual elements or limited structural systems. Designers
need to consider the ongoing effects in the damaged structure, under the influence
of gravity.
12
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.8 SYSTEMS APPROACH
In our efforts to enhance the blast resistance of a facility, we need to remain cognizant about how our designs affect the performance and viability of the facility
for nonblast events. As is always the case, there are competing goals and influences in the design of a facility, and those factors need to be balanced to achieve
the most satisfactory end product.
Consider the conflicts between the structural performance preferred for
seismic events and that preferred for explosions. One important goal in seismic
design is to force failures to occur in beams before columns, so that the
load-carrying capacity of columns is preserved even when the earthquake
induces damage. This is accomplished by detailing connections between beams
and columns so that plastic moments occur in the beams before the columns.
This is the “strong column, weak beam” approach.
Consultants designing for blast often provide for the possibility that a column
will be severely damaged by an explosion, in spite of our best efforts at prevention. When consultants assume that a column has lost its strength, they must develop alternative load paths to prevent a collapse from progressing from the initially damaged column through the structure. One form of alternative resistance
involves making beams strong and ductile enough to span over the area of damage, thereby redistributing the load on the damaged column to adjacent columns.
This requires strong beams which, if implemented without consideration of seismic response, can run counter to philosophies for robust seismic resistance.
Designers working to enhance blast resistance must also consider occupant
egress and the needs of emergency responders. Blast resistance invariably includes fenestration with blast-resistant glass. By definition, such glass is difficult
to break. Firefighters will need to use special tools and engage unusual tactics to
fight a fire in a building that is difficult to enter and vent, and that has features
that inhibit extraction of trapped occupants. Designers might need to compensate
for blast-resistance features or enhance fire resistance.
Distance is the single most important asset to a structural engineer designing
for blast resistance. The farther the explosion is from the structure, the lower
are the effects that the structure must resist. Further, there often is merit to the
construction of blast walls or line-of-sight barriers to add protection to a facility.
However, the need to create an impenetrable perimeter, and the temptation to
make it one that effectively hides the facility, can detract from the function of
the facility.
First, there is the dilemma caused by features that are intended to keep aggressors away from a building, but that also block lines of sight to the building in
the process. While such features add security, they also provide opportunities for
the aggressors to effectively hide from observers in and around the building. A
slowly developing assault may be more difficult to detect if the perimeter cannot
be monitored effectively.
Next, there is the potential impact on the quality of life for occupants of
buildings that have very robust defenses. Imposing perimeters and minimized
fenestration display the robustness and the fortresslike design intent. While
INFORMATION SENSITIVITY
13
this might be perceived as an asset for what it says to the aggressor, it also
communicates a sobering message to occupants and welcomed visitors. There
has to be a balance between the means to provide the necessary resistance
and the architectural and functional goals of the facility. Aesthetics need
consideration for most facilities.
Overall security design needs to properly balance the efforts applied to the
defense against a variety of threats. It is unsatisfactory to provide a very robust
design to resist blast if the real threat to a facility is through the mechanical system. Clients will be unhappy if security protocols address perceived threats (e.g.,
outside aggressors detonating bombs near a building), but fail to prevent real
threats (e.g., disgruntled employees intent on committing sabotage or violence
inside the facility). Any overall security evaluation needs to consider all perceived threats and provide guidance that will allow clients to determine where
best to apply their efforts to maximize their benefits. In many cases, a robust
resistance to an explosion threat will not be the best expenditure of funds.
Given a security design developed for the spectrum of potential threats to a
facility, owners sometimes face costs that exceed their means. When this occurs,
and for facilities that risk assessments show to be at relatively low risk, owners
must make decisions. Sometimes they instruct consultants to design to a particular cost, representing the amount that the owner can commit to the added security
to be provided to the facility. In these instances, consultants must identify priorities that address the most likely threats and provide the greatest protection for
the limited funds. When this happens, the consultants must explain to the owners
the limitations of the options so that the owners can make educated decisions.
1.9 INFORMATION SENSITIVITY
When blast-resistant designs are for the security and safety of a facility in response to a threat of a malevolent attack, information about the assumed size
and location of an explosion should be kept confidential. This information could
be useful to an aggressor because it can reveal a strategy to overwhelm the designed defenses.
The common practice of specifying the design loads on drawings should not
include a specific statement about the assumptions for blast loading when facility
security is at issue. Potentially public communications among members of the
design team and between the design team and the owner should avoid revelations
about the design-base explosion.
In most cases, the design assumptions for accidental explosions are not sensitive. Precautions about security-related confidentiality usually do not apply,
and customary processes for documenting the design bases may be followed.
In addition, there might be legal requirements or other circumstances that dictate the documentation of otherwise sensitive information. As always, designers
will need to comply with the law and to work with stakeholders in the design
of a facility to contain the unnecessary dissemination of information that could
potentially be misused.
14
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.10 SUMMARY
As consultants in the building design industry have been drawn into the matter of
blast-resistant design, many have been handicapped by lack of familiarity with
the blast environment, including not knowing how to determine loads for design,
or with proper approaches for structural design. Consultants often anticipate that
they will be able to provide effective designs by following approaches common
in building design when blast is not an issue. Unfortunately, consultants expecting to apply their familiar approaches usually are proceeding along an improper
path.
An explosion is a violent thermochemical event. It involves supersonic detonation of the explosive material, violently expanding hot gases, and radiation of a
shock front that has peak pressures that are orders of magnitude higher than those
that buildings normally experience under any other loadings. Designers hoping
to solve the blast problem by designing for a quasi-static pressure are likely to
be very conservative, at best, but more probably will simply be wrong.
Designers need to understand that the magnitudes of the pressures that an explosion imparts to a structure are highly dependent on the nature of the explosive
material, the shape and casing of the device, the size and range of the explosion, the angle of incidence between the advancing shock front and the impacted
surface, the presence of nearby surfaces that restrict the expansion of hot gases
or that reflect pressure fronts, and the robustness of the impacted surface itself.
Designers also need to understand that the durations of the pressures induced by
an external explosion generally are extremely short compared to the durations
of other loads and compared to natural periods of structures. Further, there is
interplay between blast pressure magnitudes and durations, which is a function
of distance from the detonation point, among other factors.
Designing for the very high peak pressures and short durations of blast loadings requires applications of principles of dynamic response. Accurate prediction
of the peak response of a building will require the designer to analyze dynamic
properties of the structure, and apply approaches that respect dynamic behavior.
Further, most cost-efficient designs rely on deformation far beyond elastic limits
to dissipate energy. Hence, many of the assumptions designers normally make
when designing for loads other than blast do not apply when designing for blast
resistance.
Consultants engaged in the design for blast resistance need to be qualified
by education, training, and experience to properly determine the effects of an
explosion on a structure. They must have specialized expertise in blast characterization, structural dynamics, nonlinear behavior, and numerical modeling of
structures. Blast resistance designers must be licensed design professionals who
are knowledgeable in the principles of structural dynamics and experienced with
their proper application in predicting the response of elements and systems to
the types of loadings that result from an explosion, or they must work under
the direct supervision of licensed professionals with appropriate training and
experience.
REFERENCES
15
The present practice for blast-resistant design employs many approximations
and, in many aspects, relies on incomplete understanding of the blast environment and structural behavior. While available approaches serve the public by
increasing the ability of our structures to resist the effects of explosions, these
conventional approaches generally are ill suited to provide a clear understanding
of the post-blast condition of the structure. Consultants providing blast-resistant
design need to understand the limitations of the tools they apply, and provide
clients with appropriate explanations of the assumptions, risks, and expectations
for the performance of blast-resistant structures. In many cases, those explanations need to make clear that the performance of the structure and the safety of
individuals inside the protected spaces are not guaranteed.
REFERENCES
American Concrete Institute. 2005. Building Code Provisions for Structural Concrete
and Commentary (ACI 318-05). Farmington Hills, MI: American Concrete Institute.
American Institute of Steel Construction. 2005. Specifications for Structural Steel Buildings. Chicago, IL: American Institute of Steel Construction, Inc.
American Society of Civil Engineers. 1999. Structural Design for Physical Security:
State of the Practice. Reston, VA: American Society of Civil Engineers.
. 2005. Minimum Design Loads for Buildings and Other Structures (ASCE/SEI
7-05). Reston, VA: American Society of Civil Engineers.
Bangash, M. Y. H. 1993. Impact and Explosion: Analysis and Design. Boca Raton, FL:
CRC Press, Inc.
Center for Chemical Process Safety. 1996. Guidelines for Evaluating Process Plant
Buildings for External Explosions and Fires. New York, NY: American Institute of
Chemical Engineers Center for Chemical Process Safety.
Federal Emergency Management Agency. 1996. The Oklahoma City Bombing: Improving Building Performance Through Multi-Hazard Mitigation (FEMA 227).
Washington, DC: Federal Emergency Management Agency.
. 2003. Reference Manual to Mitigate Potential Terrorist Attacks Against Buildings (FEMA 426). Washington, DC: Federal Emergency Management Agency, Department of Homeland Security.
General Services Administration. 2003. Facilities Standards for Public Buildings Service
(P100). Washington, DC: General Services Administration.
Hinman E. E. and D. J. Hammond. 1997. Lessons from the Oklahoma City Bombing:
Defensive Design Techniques. Reston, VA: American Society of Civil Engineers Press.
Mays G. C. and P. D. Smith. 1995. Blast Effects on Buildings. London: Thomas Telford
Publications.
National Research Council. 1995. Protecting Buildings from Bomb Damage: Transfer of Blast-Effects Mitigation Technologies from Military to Civilian Applications.
Washington, DC: National Research Council, National Academy Press.
Smith P. D. and J. G. Hetherington. 1994. Blast and Ballistic Loading of Structures.
Oxford: Butterworth Heinemann.
16
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
U.S. Department of Defense. 2002a. DoD Minimum Antiterrorist Standards for Buildings
(UFC 4-010-01). Washington, DC: United States Department of Defense.
. 2002b. Design and Analysis of Hardened Structures to Conventional Weapons
Effects (DAHS-CWE) (UFC 3-340-01) (TM 5-855-1). Washington, DC: United States
Department of Defense.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002), Washington, DC: United States Department of Defense.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
2
Design Considerations
Robert Ducibella and James Cunningham
2.1 INTRODUCTION
On April 19, 1995, a Ryder truck containing about 5,000 pounds of ammonium nitrate fertilizer and nitromethane exploded in front of the Alfred P.
Murrah Federal Building in Oklahoma City, Oklahoma. The blast collapsed a
third of the building, gouged out a crater 30 feet wide by 8 feet deep, and destroyed or damaged 324 buildings in a 16-block radius. The bomb claimed 168
confirmed dead.
On September 11, 2001, terrorists crashed two commercial jets into the
World Trade Center twin towers. Another hijacked flight slammed into the
Pentagon, while a fourth was forced down by passengers and crashed in a field
near Shanksville, Pennsylvania. The total dead and missing numbered 2,992:
2,749 in New York City, 184 at the Pentagon, 40 in Pennsylvania, and the
19 hijackers.
More than any other acts of domestic or international terrorism, these two
attacks have forever changed the American building-sciences community’s relationship to the society it serves. Since the first skyscrapers appeared in the late
nineteenth century, Americans have come to expect commercial structures of
exceptional beauty and functionality. After Oklahoma City and September 11,
many ordinary citizens also assume that new buildings are designed to protect
people during explosions as well as other natural or man-made disasters. With
few exceptions, however, significant movement toward achieving the recent imperative of both commercial utility and explosive blast resistance is a work in
progress.
Design professionals have gained enormous experience with plans and models
that anticipate structural responses to gravity, wind, and seismic loads. Preventing or curtailing random acts of terrorism by identifying their probability of occurrence and potential consequences, however, falls outside the general practice
of structural engineering.
This chapter proposes an innovative and largely untapped approach to blastresistant-building security design, a new paradigm in which senior individuals
who have a breadth and depth of experience in the areas of site planning, civil,
17
18
DESIGN CONSIDERATIONS
structural, mechanical, electrical, fire protection, and vertical transportation engineering; architecture; code and egress consulting; site planning; and security
engineering collaborate on a total blast-resistant building security design. This
team approach should take into consideration security and antiterrorist strategies
that fundamentally affect site selection and building design. In considering the
design of blast-resistant buildings, the design professionals must partner to
become an effective security design team.
In more traditional building security design efforts, security professionals simply present the design team with the results of their risk assessment, and the
security planning component is assumed to be largely finished. For reasons expanded upon hereinafter, the authors urge instead that the design team collaborate closely with security professionals and security engineers throughout the
entire design process. In the iterative process of designing a blast-resistant facility, the architectural team should become a security design team, supported by
homeland defense, intelligence, security, law enforcement, and blast consultant
experts. Therefore, the term security design team as used in this chapter means
the multidisciplinary building sciences/security/explosives experts group of professionals described above.
It is a concept capable of implementation and proven to work in a wide range
of facility types and locations where occupants, assets, and business missions are
deemed worthy of protection.
2.2 A NEW PARADIGM FOR DESIGNING BLAST-RESISTANT
BUILDINGS, VENUES, AND SITES
The following paragraphs describe a structured framework for threat and vulnerability data gathering and for risk assessment. Security concepts such as design basis threat, consequence management, functional redundancy, building location, and critical-functions dispersal are explored. A brief checklist of security
design considerations is presented, and the reader is introduced to the design
principles and guidelines that are expanded upon in the handbook’s subsequent
chapters.
The suggested risk assessment model for blast protection has six parts:
1. A threat identification and rating, which is the security design team’s analysis of what terrorists and criminals can do to the target.
2. An asset value assessment, which represents how much the project’s people
and physical assets are worth and what the responsible parties will do (and
pay) to protect them.
3. A vulnerability assessment, which represents the attractiveness of the target, and areas of potential weakness and/or avenues of compromise.
4. A site-specific risk assessment, which is the product of these three studies.
A credible site-specific risk assessment is the single most critical factor
A NEW PARADIGM FOR DESIGNING BLAST-RESISTANT BUILDINGS
Threat
Identification
and Rating
(Step 1)
Vulnerability
Assessment
(Step 3)
Asset Value
Assessment
(Step 2)
19
Benefits Analysis
How mitigation options
change the vulnerability
and ultimately the risk.
Risk
Assessment
(Step 4)
Consider
Mitigation
Options
(Step 5)
Risk
Management
Decision
Cost Analysis
How mitigation options affect
the asset’s criticality
and ultimately the risk.
Figure 2.1 Risk Assessment Process (Adapted from FEMA 452)
of the security blast-design process; it is the basis and rationale for ensuring that the protective design strategies are incorporated in the multidisciplinary security/explosives/building sciences team approach that is stressed
throughout this chapter.
5. Mitigation options, based on the risk assessment as a foundation.
6. Risk management decisions, driven by the mitigation options and informed
by the available project resources.
Figure 2.1 shows the five steps leading to the risk management decisions, and
their interaction with each other.
To be useful in influencing the design, the risk assessment should be completed early, preferably during the preliminary planning and conceptual design
phase, but certainly no later than the completion of initial design documents.
Extensive unclassified information about explosive events is available from
the FBI; Department of State; Department of Defense (DoD); Department of
Homeland Security (DHS); Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF); the U.S. Armed Forces; and other U.S. agencies. The authors urge
the design team to obtain and use this information. Since this information is being constantly updated, we advocate that you use an Internet search engine such
as Google, Mozilla Foxfire, or Ask.com to conduct your own research, thereby
ensuring that your information is both current and relevant. For example, a quick
search on the terms “Department of Defense explosive events” results in over
2 million hits while “FBI explosive events” returns over 600,000 citations, and
“ATF explosive events” creates over 200,000 references.
20
DESIGN CONSIDERATIONS
There are numerous how-to guides that lay out systematic approaches
to secure facility planning, design, construction, and operation. These various methodologies—the Department of Defense CARVER process, Sandia’s
RAMPARTTM software tool, the National Institute of Standards and Technology (NIST)’s CET, and the Federal Emergency Management Agency (FEMA)
building security series are among scores of candidates1 —are all based on a
process of threat assessment, vulnerability assessment, and risk analysis.
The authors have adapted and somewhat modified FEMA’s Risk Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks Against Buildings
(FEMA 4522 ) and FEMA’s Reference Manual to Mitigate Potential Terrorist
Attacks Against Buildings (FEMA 426) as their risk assessment models to
present the risk management process. Admittedly, FEMA 452 is a somewhat arbitrary choice of exemplar, although it is becoming a de facto industry standard
among security professionals. There are many alternative methods—an Internet
search of the term “risk assessment for buildings” resulted in over 3.7 million
hits—but no matter which risk assessment version the design team selects, the
methodology should be:
r Specifically written for the building sciences community of architects, engineers, and professionals who design not only high-security government
facilities but private-sector structures as well.
r Intended to serve as a multi-hazard assessment tool of a building and its
site, but readily adaptable to focusing closely on the explosive threat.
r Organized with numerous checklists, tables, and memory aids that will assist the design team in determining threats, risks, vulnerabilities, and mitigation options.
r Proven effective through years of extensive use in real-world threat and vulnerability assessments.
1
The DoD’s CARVER risk assessment process is a mnemonic rather than a model. First developed for the U.S. Special Forces in Viet Nam to target enemy installations, CARVER
stands for Criticality, Accessibility, Recognizability, Vulnerability, Effect, and Recoverability.
CARVER has recently become hard to find on the Internet but is widely available through
the DoD, Homeland Security, ASIS International, or law enforcement agencies, among others. Sandia’s Risk Assessment Method—Property Analysis and Ranking Tool (RAMPART(tm))
is a software-based methodology for assessing the potential risks of terrorism, natural
disasters, and crime to buildings, particularly U.S. government facilities. Read more at
http://ipal.sandia.gov/ip details.php?ip=4420. NIST’s Cost-Effectiveness Tool for Capital Asset
Protection (CET) is a software-based risk assessment tool that building owners and managers can
use to protect assets against terrorist threats. CET is available without cost on the NIST Web site.
See http://www2.bfrl.nist.gov/software/CET/CET 4 0 UserManualNISTIR 7524.pdf.
2
FEMA, located within the Department of Homeland Security, is the U.S. government agency
tasked with disaster mitigation, preparedness response, and recovery planning. FEMA 452: Risk
Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks may be downloaded from
www.fema.gov/plan/prevent/rms/rmsp452.shtm. FEMA 452 is a companion reference to the Reference Manual to Mitigate Potential Attacks Against Buildings (FEMA 426) and the Building Design
for Homeland Security Training Course (FEMA E155). This document is also a useful companion
to the Primer for Design of Commercial Office Buildings to Mitigate Terrorist Attacks (FEMA 427).
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
21
r Available in unclassified form and preferably at little or no cost. FEMA
452, FEMA 426, and their companion manuals, for example, are available
for free from http://www.fema.gov/plan/prevent/rms/rmsp452.shtm.
Increasingly, the building sciences community is being challenged to incorporate high levels of security into the design of facilities, sites, and venues
that do not yet exist. While the risk assessment tools spreading throughout
the law enforcement, public safety, security, and building sciences communities can be extremely useful, almost all of them are geared to improving
the security of sites and structures that are already standing. Consequently,
it is essential that the risk assessment model that is selected has also been
used to assess future buildings and venues, and to create security plans for
virtual sites.
Traditionally, the building sciences community has prepared for natural disasters by following prescriptive building codes supported by well-established and
tested reference standards, regulations, inspections, and assessment techniques.
Many man-made hazards such as toxic industrial chemicals storage, and numerous societal goals such as life safety, have been similarly addressed. The building
regulation system, however, has only just begun to deal with the terrorist threat.
In the absence of high-quality regulatory guidance, the design team must fall
back on its own resources and expertise—always remembering that the nature of
the potential threat and the desired level of protection are equally important and
inseparable design considerations.
2.3 A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
Experts are quick to point out that terrorists almost invariably seek publicity
and sometimes monetary reward or political gain as well. It should be emphasized, though, that terrorism has a powerful appeal to many of the marginalized people of the world. Throughout human history, asymmetric warfare—in
this case, terrorism—has had an undeniable allure to some and the sympathy of
many more. Because it can be spectacularly effective, terrorism will be around
for the foreseeable future, hence the legitimate concern for designs to mitigate
its effects.
As George Santayana famously said, “Those who cannot remember the past
are condemned to repeat it.” What follows, therefore, is a brief survey of broad
trends in domestic and international terrorism, for the purpose of planning future
security measures in site selection and facility design by learning from history
(Santayana 1905, 284).
2.3.1 Terrorists’ Use of Explosives
Explosives continue to be the terrorist’s preferred weapon, since they are destructive, relatively easy to obtain or fabricate, and still comparatively easy to
move surreptitiously on the ground and by sea. Terrorists are also well aware
22
DESIGN CONSIDERATIONS
that explosives produce fear in the general population far beyond the geographical location of their intended target.
2.3.2 Vehicle-Borne Improvised Explosive Devices
The following case studies include just some of the vehicle-borne explosive devices that have been used in the past quarter-century.
April 18, 1983—The modern era of vehicle-borne improvised explosive devices dates from April 1983, when a massive truck bomb destroyed the
U.S. Embassy in Beirut. The blast killed 63 people, including 17 Americans. The attack was carried out by a suicide bomber driving a van, reportedly stolen from the embassy in June 1982, carrying 2,000 pounds of
explosives.
October 23, 1983—A Shiite suicide bomber crashed his truck into the lobby
of the U.S. Marine headquarters building at the Beirut airport. The explosion, the equivalent of 12,000 pounds of trinitrotoluene (TNT) and alleged
to be the largest truck bomb in history, leveled the four-story cinderblock
building, killing 241 servicemen and injuring 60. Minutes later a second
truck bomb killed 58 French paratroopers in their barracks in West Beirut.
September 20, 1984—A van loaded with an estimated 400 pounds of explosives swerved around several barricades and U.S. soldiers and penetrated
the relocated U.S. Embassy annex compound in East Beirut. The suicide
bomb exploded 30 feet from the building, killing 11, including 2 U.S. servicemen, and injuring 58.
February 26, 1993—World Trade Center’s Tower One’s underground parking garage was rocked by a powerful explosion. The blast killed 6 people
and injured at least 1,040. The 1,310-pound bomb was made of urea nitrate
pellets, nitroglycerin, sulfuric acid, aluminum azide, magnesium azide, and
bottled hydrogen—all ordinary, commercially available materials. The device, delivered in a yellow Ryder rental van, tore a crater 100 feet wide
through four sublevels of reinforced concrete.
April 19, 1995—Timothy McVeigh detonated an ammonium nitrate/fuel oil
bomb in front of the Alfred P. Murrah Federal Building in downtown
Oklahoma City, Oklahoma.
August 7, 1998—Two truck bombs exploded almost simultaneously at U.S.
embassies in two East African capitals, killing 213 people in Nairobi and
11 more in Dar es Salaam. Some 4,500 individuals, principally Kenyans
and Tanzanians, were injured. In May 2001, four men connected with
al-Qaeda, two of whom had received training at al-Qaeda camps inside
Afghanistan, were convicted of the killings and sentenced to life in prison.
A federal grand jury has indicted 22 men, including Osama bin Laden, in
connection with the attacks.
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
23
December 14, 1999—An apparent plot to bomb the Los Angeles airport was
disrupted when Ahmed Ressam, a Canadian of Algerian background, was
arrested at a United States–Canada vehicle border crossing in Washington
State. Ressam had nitroglycerin and four timing devices concealed in his
spare-tire well.
September 11, 2001—Al-Qaeda terrorists crashed two commercial jets into
the World Trade Center twin towers, and another hijacked flight slammed
into the Pentagon, while a fourth was forced down by passengers and
crashed in a field near Shanksville, Pennsylvania.
June 14, 2002—A powerful fertilizer bomb blew a gaping hole in a wall outside the heavily guarded U.S. Consulate in Karachi, Pakistan. The truck
bomb, driven by a suicide bomber, killed 12 and injured 51, all Pakistanis.
October 12, 2002—A bomb hidden in a backpack ripped through Paddy’s Bar
on the Indonesian island of Bali. The device was small and crude, but it
killed the backpack owner, likely a suicide operative. The bar’s occupants,
some of them injured, immediately ran into the street. Fifteen seconds later,
a second much more powerful bomb—estimated at slightly more than a ton
of ammonium nitrate—concealed in a white Mitsubishi van was detonated
by remote control in front of the Sari Club. This blast killed 202 and injured
another 209.
May 12, 2003—Attackers shot security guards and forced their way into three
housing compounds for foreigners in the Saudi capital of Riyadh. The terrorists then set off seven simultaneous car bombs, which killed 34 people,
including 8 Americans and 9 Saudi suicide attackers, and wounded almost
200 more. The facades of four- and five-story buildings were sheared off.
One explosion left a crater 20 feet across, while several cars and six or
seven single-family homes within 50 yards of the blast were destroyed.
August 5, 2003—A powerful car bomb rocked the JW Marriott hotel in central Jakarta, Indonesia, killing 12 people and injuring 150. Police believe
the suicide bomb, which severely damaged the American-run hotel, was
concealed inside a Toyota car parked outside the hotel lobby. The terrorist
group Jemaah Islamiyah is believed responsible.
August 19, 2003—Sergio Vieira de Mello, the U.N. special representative in
Iraq, and at least 16 others died in a suicide truck-bomb explosion that
ripped through the organization’s Baghdad headquarters. The bomb-laden
truck smashed through a wire fence and exploded beneath the windows of
de Mello’s office in the Canal Hotel in the late afternoon, destroying the
building and shattering glass a half mile away. The concrete truck was said
to have been chosen as a Trojan horse because of all the construction work
going on in the area.
November 8, 2003—Seventeen people, including 5 children, were killed and
more than 100 were wounded in an armed raid and suicide car-bomb attack
on a residential compound in Riyadh, the Saudi Arabian capital. The huge
24
DESIGN CONSIDERATIONS
blast from an explosives-laden vehicle came after gunmen tried to break
into the compound and exchanged fire with security guards. The bomb
could be heard across the capital, and Riyadh residents who lived miles
away said they felt their buildings shake.
March 11, 2004—The routine of a morning rush hour in Madrid, Spain, was
shattered when 10 simultaneous explosions occurred aboard four commuter trains. The attacks killed 191 and wounded more than 1,800. Thirteen improvised explosive devices concealed in backpacks were used, but
three failed to detonate and were recovered by the authorities.
October 8, 2004—The Hilton hotel in the Red Sea resort of Taba, Egypt, was
rocked by a large explosion that killed 29 people and injured at least 120
more. The powerful blast was caused by two separate suicide car bombs,
each containing about 200 kilograms of explosives, which crashed into the
lobby before detonating simultaneously.
March 2, 2006—A suicide car bomb killed 4 people and injured 50 others in a
parking lot adjacent to the Marriott hotel in Karachi, Pakistan. The hotel is
about 30 yards from the U.S. Consulate. Among the dead was David Foy,
an American diplomat believed to have been the target of the attack. Police
said they believed the bomber rammed a car packed with high explosives
into the diplomat’s vehicle. The force of the blast, the most powerful of its
kind in Karachi to date, lifted the victim’s vehicle into the air, hurled it over
a 7-foot wall, and left a crater 6 feet deep.
June 2 and 3, 2006—Canadian security and intelligence services arrested 12
men and 5 youths on charges of plotting to bomb targets in southern Ontario, including Parliament. The terrorist ring had attempted to procure 3
tons of ammonium nitrate.
June 29 and 30, 2007—A car driven by suspected al-Qaeda sympathizers
rammed into the main terminal of Glasgow International Airport, setting
off a large blaze when the explosive devices caught fire instead of detonating. The previous evening in London’s West End, an attempt to set off
two car bombs was unsuccessful when the mobile phone triggering devices
failed.
2.3.3 Person-Borne Improvised Explosive Devices
Person-borne improvised explosive devices (PBIEDs) are usually less powerful
than vehicle-borne improvised explosive devices, but they are potentially more
deadly because they are usually detonated very close to their intended victims.
In general, PBIEDs are targeted against civilian populations or at military personnel in civilian settings such as bus stops. The act usually has a political (not
military) purpose, since the intention is to kill, maim, and terrorize a civilian population, and the target areas have few or no security measures in place to prevent
bombings. For example, the so-called Second Intifada in Israel and the occupied
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
25
territories, which occurred between October 2000 and October 2006, saw 267
clearly identified suicide bomber attacks, most of which involved person-borne
explosives. More recent attacks in Iraq and Afghanistan continue this deadly
trend.
PBIEDs are generally delivered to the target in the following ways.
r The explosives are transported in a package or backpack, and the bomber
leaves the area before the explosion. The Madrid commuter attacks of 2004
and the Centennial Olympic Park bombing3 in Atlanta, Georgia, during the
1996 Summer Olympics illustrate this technique.
r The PBIEDs are worn by the attacker, who expects to die in the explosion.
(Suicide bombers, including female terrorists, are increasingly used to carry
out such attacks.) The Amman, Jordan, hotel bombings of November 2005
were suicide attacks, as were the numerous Intifada-inspired explosive-vest
attacks in Israel following the collapse of the Camp David II summit in
2000.
r The person-borne bomb may not always, in fact, be carried in the traditional
sense of the word, but may instead be placed in a wheelchair, a wheeled
computer case, or a small cart pulled by the aggressor. This delivery mechanism has the advantages of (1) depriving the security forces of the visual
cues presented by a person with a bulky object under his or her clothes and
(2) allowing the bomber to deliver more explosives. Consequently, its potential explosives design basis threat (DBT) value—a description of a specific
explosive agent, a quantitative value, and a mode of attack that has a degree
of likelihood of occurring—is significantly greater than that of body-worn
devices. Mitigating the threat of a PBIED is essential in facility designs,
as its successful delivery and detonation have the ability to invoke mass
casualty, disproportionate damage, and progressive collapse.
r The person who delivers the bomb may be unaware of his or her role in
the act of terrorism—for example, convicted Unabomber Ted Kaczynski’s
use of postal delivery workers. Express couriers, common freight carriers,
or other similar second- and third-party delivery agencies such as FedEx or
UPS might also be used. These bombs differ from the usual PBIED threat,
as they arrive cloaked in the legitimacy associated with an expected delivery mechanism and deny security personnel the opportunity to profile the
actual bomber. Using this delivery method, the explosive DBT value could
be equal to the carrier’s parcel weight and size management thresholds.
2.3.4 Locally Available Explosives
In assessing a threat, perhaps the most important question to answer is “Can the
enemy obtain weapons?” The recent wave of terrorist bombings and bomb plots
3
See http://en.wikipedia.org/wiki/Centennial Olympic Park bombing for details.
26
DESIGN CONSIDERATIONS
around the world clearly answers that question, but the apparent ease with which
powerful explosives are available is worth discussing.
Shortly after the Taliban retreated from Kabul in November 2001, a search
of an al-Qaeda safe house uncovered a bomb-making training center. Among
the documents seized were a detailed explosives instruction manual, a table of
explosive mixtures classified by strength, and a table comparing the power of
potential detonators such as acetone peroxide (Boettcher and Arnesen 2002).
Shoe bomber Richard Reed used triacetone triperoxide (TATP) in his unsuccessful December 2001 attempt to down an American Airlines flight en route
from Paris to Miami. In July 2005, however, the four London bus and subway bombers detonated homemade acetone peroxide explosives with deadly
effect.
The al-Qaeda safe house also yielded handwritten lists of formulas that included instructions on how to make RDX and a version of C-4, the powerful
military explosive used against the USS Cole in December 2000. The documents
suggest that al-Qaeda was developing its own variant of C-4, known commercially as Semtex, for use in a wide variety of bombs. Creating explosives from
the base chemicals, the training manuals noted, would avoid the security problems inherent in finding a supplier and in transporting the contraband into the
target country.
The ability to buy explosives components locally “gives more latitude, more
autonomy and possibly some degree of elusiveness,” according to Tony Villa,
an explosives expert who has worked extensively for the U.S. government
(Boettcher and Arnesen 2002). Al-Qaeda is “not just a bunch of guys climbing along some jungle gym and going through tunnels and shooting their guns
in the air,” notes David Albright, a nuclear weapons design and proliferation expert (Boettcher and Arnesen 2002). “These are people who are thinking through
problems in how to cause destruction.”
Al-Qaeda videotapes, discovered in Afghanistan after the Taliban’s defeat,
also show a level of bomb-making skill that could allow terrorists to arrive unarmed in the target city and to easily buy and mix the ingredients of a highexplosive device. One of the captured videotapes, a detailed how-to guide,
demonstrates the production from scratch of high-quality TNT and other bomb
components using easily obtained materials. The videotaped demonstration reinforces the written bomb manuals but also makes the training much more
effective.
“What we did see,” Albright said, “is that when we compared this information
on high explosives to the Internet that these [training manuals] are much more
polished. They really did work with these formulas, tested these formulas, and
developed a procedure of making these high explosives that led to effective high
explosives in a safe manner (Boettcher and Arnesen 2002).”
“The overarching point here is that they can pick any venue or target city,
arrive in that city and, based on the video tapes, construct a bomb using common
materials,” Villa explained (Robertson 2002).
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
27
2.3.5 Some Counterterrorism Considerations
Counterterrorism officials have drawn several lessons from these and many similar attacks:
1. Terrorists have always stuck to the operational maxim that what is simple
is best, so the experts believe that increased security measures have pushed
terrorists toward softer targets and easier-to-deliver but deadly vehicleborne bombs.
2. The knowledge needed to create bombs and biological/chemical weapons
is available in bookstores, in widely circulated training manuals, on
CD-ROM, and on the Internet. Armed with this information, even an amateur terrorist can buy enough commercially available chemicals to make
his or her contribution to global terrorism.
3. As Oklahoma City bomber Timothy McVeigh and the four July 2005
London transit jihadists have demonstrated, even novice terrorists can
purchase bomb-making components and mix the ingredients to create a
high-explosive device. For example, terrorists have combined acetone, hydrogen peroxide, and a strong acid such as hydrochloric or sulfuric acid
to make triacetone triperoxide (TATP), a powerful and highly unstable
explosive.
4. In the brief sample of bombing case studies above, bombers detonated their
weapons simultaneously or nearly simultaneously 29 times, including the
7 synchronized blasts in Riyadh and 10 in Madrid.
5. Explosive charges greater than 2,000 pounds were used in nine of the
bombings described above. Weapons weighing less than 1,000 pounds appear to have been used in eight cases, while the remaining seven explosive
devices were described in terms such as “powerful” and “massive” without identifying the actual amount of explosives. This brief sample suggests
that the continued use of large bombs carried in vans, SUVs, trucks, or
passenger vehicles reinforced for this purpose is likely.
6. Suicide targets were often at hotels, in restaurants, and on public
transportation—places where civilians had gathered.
7. Terrorists have planned phased explosions to target emergency responders,
civilians who have not yet been evacuated, and bystanders.
8. Aggressive military, intelligence, and law enforcement measures since
September 2001 have disrupted the terrorists’ communications and logistical support. These worldwide activities have degraded the radical jihadists’
ability to mount operations as sophisticated and coordinated as the hijackings of September 11 and the 1998 bombing of the embassies in East
Africa. As the message of jihad reaches more disaffected young men, however, the bombings may be smaller in scale but more frequent.
28
DESIGN CONSIDERATIONS
9. The terrorists will opt for continual low-level bombings, therefore, as part
of an ongoing campaign to inflict some “revenge”—less dramatic and less
traumatic than the twin towers—on the enemy and to prove their relevancy.
10. The terrorists sought out and exploited gaps where security was either ineffective or lacking—most notably in the case of the U.S. commercial aviation industry on September 11, 2001.
11. Today’s terrorists harbor no compunctions about killing or injuring innocent civilians.
2.4 COLLABORATING TO ANALYZE RISK
The first risk assessment probably occurred when a prehistoric couple heard
strange noises outside their cave; in the intervening millennia, thousands of risk
assessment formats have been developed. Today, in response to private-sector
concerns, the security community’s post-9/11 threat and vulnerability evaluations place far more emphasis on terrorism and criminal violence than they did
even a decade ago.
A comprehensive risk assessment for blast-protection design involves
close collaboration among city planners, architects, engineers, blast consultant subject matter experts, and law enforcement security professionals. In
this group effort, collaborating professionals assess and select the security
measures needed to detect, deter, prevent, defeat, mitigate, or recover from
terrorists’ bomb attacks.
This section introduces the analytical and decision-making steps that are
widely used today to identify the critical assets requiring protection, determine
the explosive threats, assess the vulnerabilities, and create mitigating measures.
It begins with a thorough but flexible five-step methodology for preparing and
executing a risk assessment.
2.4.1 Step 1—Threat Identification and Rating
The assessment process begins with a detailed understanding of the natural and
man-made threats that could realistically occur in the geographic area in question
and therefore pose a risk to the building or site being designed. To best prepare
the design team to anticipate the broadest range of contingencies, the threat identification and rating step calls for the following four tasks.
Task 1-1. Identifying the Threats. In blast-protection terms, a threat is a potential event culminating in an explosion that damages or destroys a protected
facility or its inhabitants.
Task 1-2. Collecting Information on Potential Attacks. Collecting information
for an explosives-oriented threat assessment requires expertise in weapons and
demolitions as well as in terrorist history, tactics, training, and targeting. An
informed estimate of the aggressors’ motivation and intentions against the future
COLLABORATING TO ANALYZE RISK
29
facility and its future occupants is also essential. The following list of questions
is a quick overview of the kinds of information the team needs to gather:
r What domestic or international terrorist groups are the potential aggressors,
and what are their capabilities?
r What has been expressed or can be inferred about their motivation and intentions against the clients?
r How skilled with demolitions are they? Are bomb-making materials available to them?
r Does the project have an economic, cultural, or symbolic significance that
terrorists could exploit to justify their acts and to gain publicity?
Many informative checklists and memory aids are available for collecting
these data, and several of the best have already been cited. The design team’s
security members should be well versed in at least one of these tools. They
should also have information sources among state and federal law enforcement
and homeland security agencies—if not by virtue of existing relationships, then
certainly through contacts and their own credibility. Neighboring building owners, operators, and tenants can also be a valuable data source.
Task 1-3. Determining the Design Basis Threat for Explosives. The Department of Defense defines the design basis threat (DBT) as the hazard against
which a building must be protected and upon which the protective system’s design is therefore based.4 In other words, the type and size of weapon the building
and site must be designed to withstand is the explosive DBT.
To calculate an explosive DBT, the team must first estimate the potential for an
attack by identifying the potential terrorists’ tactics and the tools, weapons, and
explosives they could employ. Terrorists have used explosives hidden in moving
and stationary vehicles and have left behind explosives that were hand-carried
to the target. Mail bombs and supply bombs—larger devices typically infiltrated
through shipping departments—and explosive devices carried by suicide terrorists are also proven techniques. For reference purposes, large-scale truck bombs
typically contain 10,000 pounds or more of TNT equivalent (all the following
bomb weights are in terms of TNT equivalent). Bombs in vehicles ranging from
small sedans to vans typically contain 500 to 4,000 pounds. Suicide bombers can
unobtrusively deliver belts ranging in size from less than 10 pounds up to about
40 pounds. Hand-carried explosives are typically on the order of 5 to 10 pounds
(Federal Emergency Management Agency 2005, 1–7).
Design documents do not normally specify the actual DBT explosive weights.
Instead, the weights are expressed in terms such as W-1 (for example, a 10-pound
bomb), W-2 (100 pounds), W-3 (1,000 pounds), and so forth. The actual values
may vary from project to project as determined appropriate by the design team.
4
This DoD definition of “threat” can be found at www.dtic.mil/doctrine/jel/doddict/data/d/01635.
html.
30
DESIGN CONSIDERATIONS
If design documents containing specific DBT explosive weights fell into terrorist
hands, they would become an extremely valuable aid to the enemy. Actual DBT
weights are revealed only on a need-to-know basis.
The willingness of suicide terrorists to turn themselves into bombs precludes
design professionals from adopting a design basis threat of “no explosives.” Consequently, when the design team is developing protective design strategies for a
site or building and its occupants, it must insist on a realistic cataloging of the
project-specific explosive DBTs. This information mining is essential if structural and architectural facility designs are to provide credible protection.
The structural engineers’ and architects’ protection strategies are based on
an estimation of the possible size and location of a detonation, coupled with a
realistic estimate of the facility’s ability to withstand an attack. Since the eventual design implementation concepts and components of a facility’s structure;
fac¸ade; evacuation, rescue, and recovery systems (ERR); and business continuity
measures are built upon this analysis, establishing quantifiable DBTs is a fundamental first step in providing the design team with a framework for designing
mitigation measures.
Task 1-4. Determining the Threat Rating. The team’s threat ratings will evaluate the probability of various blasts occurring and the likely consequences if they
do. The team should imbed probability and consequences in their consciousness;
they are two extremely important concepts that security design planning teams
far too often overlook.
Vehicle-borne bombs continue to be the terrorists’ favorite tactic, primarily
because trucks can carry potentially devastating explosive cargos. But terrorists
will exploit the weakest links in a building’s protective design; person-carried
bombs have shown that relatively small weapons can cause great damage to exposed and structurally vulnerable building interiors. This reality appears to argue for defending buildings against every conceivable explosive attack, but total
protection through design and engineering is impractical in terms of structural
hardening and benefit-to-cost ratio. A potential solution for the “protect everything” dilemma is discussed in later paragraphs, when the authors address the
complementary roles of building design, physical security, operational security,
and continual situational awareness.
Facility designs outside North America are, of course, also vulnerable to explosives threats. For work outside of North America, it is especially important to
develop a comprehensive risk assessment that is specific to the geographic area,
the project site, and the facility.
In addition to the person- and vehicle-borne IED delivery methods, the use
of standoff military ordnance, especially the use of shoulder-fired arms such
as the ubiquitous rocket-propelled grenade (RPG), remains a significant concern. Structural engineers and architects should give careful consideration to
these military-grade weapons as part of a competent protective design threatmitigating strategy. Also, major headaches can be avoided by resisting the tendency to overestimate the man-made threat. Ninja-like terrorists in attack helicopters have not appeared yet, and the two-millennium history of terrorism
COLLABORATING TO ANALYZE RISK
31
clearly demonstrates that the enemy realizes that the best route to success is very
often the easiest one.
The design team must provide project-specific information at the earliest
stages in which conceptual designs are adequately advanced. The architects’ and
designers’ timely participation in the initial risk assessment process will greatly
influence the credibility—or lack of credibility—of the project design team’s
analysis of potential vehicle and pedestrian access routes. In short, the determination of the degree to which the project site, physical plant, and personnel are at
risk from specific explosives must be informed by an interactive process of information exchange between the threat-assessment and design-team professionals.
The team, concluding its work on threat assessment, now produces a single
threat rating for the entire project. Threat ratings are commonly expressed in
a seven-level linguistic scale from very low probability to very high probability. “Very high” threats are considered credible and imminent, and “very low”
threats are deemed negligible. Between those two extremes, the spectrum includes threats of high, medium high, medium, medium low, and low probability.
The scale could be numerical instead, from 1 (very low) to 10 (very high). Some
threat-ranking formats result in both linguistic and numeric ratings. This veryhigh-to-very-low scheme is also used in ratings of asset value and vulnerability.
Figure 2.2 illustrates the use of worksheets to consolidate the design team’s
knowledge in a single, agreed-upon format. In this example, the worksheet displays the probability of various improvised and military explosives being used
in an attack and the anticipated damage as a consequence of the assault.
The authors encourage teams to use their own professional experience to question the raw scores produced by the threat-ranking scales. Applying a “Does this
make sense?” filter is wise, because rating threats is more art than science. Deferring to the client’s insistence that his facility is a likely terrorist target, for
example, puts the assessment process at risk of being driven toward the “very
high” end of the scale, resulting in overdesign. Of course, it is equally possible
to underrate the danger, and the consequences of underdesigning could be grave.
This threat identification and rating exercise creates a prioritized team assessment of the potential dangers, especially of a terrorist bombing, to the project.
2.4.2 Step 2—The Asset Value Assessment
The team will quickly need to learn as much about their building, its site, and the
surrounding area as they have discovered about the explosive threat. These assets
can be tangible, including facilities, equipment, inventory, and, of course, people.
Intangible assets include intellectual property, proprietary processes, reputation,
corporate information, and image. A concerted effort should be made to become
thoroughly acquainted with the client’s facility and its adjacent environment. The
team’s knowledge about those unique tangible and intangible assets that require
enhanced protection will be invaluable as the security design is developed.
In asset assessment, the first of four tasks is to identify the project’s layers of
defense.
32
DESIGN CONSIDERATIONS
Score
Collateral Damage at
Specified Distance
Site Population and
Capacity
Asset Accessibility
Asset Visibility and
Symbolic Value
Threat History for Buildings,
Tenants, and Environs
Knowledge and Expertise
Explosive Threats
Access to Explosives
Threat Determination Criteria
Improvised and Millitary Standoff Explosive Devices
1-lb. Mail Bomb
9
9
5-lb. Pipe Bomb
9
9
RPG or Similar Grenade/
Rocket Launcher
7
7
20-lb. Cargo Bomb
9
9
35-lb. Salchel Bomb/
Suicide Bomber
8
8
500-lb. Car Bomb
6
8
5,000-lb. Truck Bomb
4
8
20,000-lb. Truck Bomb
2
6
Natural Gas
2
8
NOTE: The threat numbers are national and are used for illustrative purposes only.
Figure 2.2 Primary Explosive Threats to Notional Urban Building (Adapted from
FEMA 426 and FEMA 452)
Task 2-1. Identifying the Layers of Defense. Layered facility protection is a
time-tested security design engineering approach. Defense-in-depth creates concentric rings of security, each increasingly difficult to penetrate, which provide
security and law enforcement officials with multiple opportunities for warning
and reaction. A layered defense complicates the terrorists’ operations, perhaps
to the point that security forces are able to interrupt the aggressors’ surveillance
or intercept them early in their attack. The obstacles presented by a layered defense may cause the terrorists to attack a less well-defended target, or could allow
building occupants time to relocate to predetermined defensive positions or find
shelter in designated safe havens during an assault.
COLLABORATING TO ANALYZE RISK
33
The first layer of defense in a typical urban setting is the surrounding
area—the sidewalks, curb lanes, streets, other buildings, and the neighborhood
itself. The structural composition of adjacent buildings can either enhance the
project’s security or increase its threat. Local law enforcement and traffic authorities are increasingly willing to support a security designer’s street and curb-lane
safety initiatives, particularly when realistic procedural measures precluding adjacent vehicle traffic lanes or parking are planned. Unobtrusive physical security devices—such as decorative anti-ram, crash-resistant street lights, benches,
planters, and trash receptacles—are found often now on sidewalks and plazas.
Embedded barricades, retractable bollards, and hydraulic barricades are also becoming more commonplace.
The exterior space between the line of vehicle or pedestrian interdiction and
the project structure is the second layer of defense. Surveillance cameras and
lighting must be part of any security design, but force protection engineered
vehicle interdiction devices and pedestrian approach restrictions can be useful
here too. Crime Prevention through Environmental Design (CPTED),5 a multidisciplinary approach to deterring criminal behavior, may also create a disincentive to terrorist surveillance and operations. ASIS International,6 the society
for professional security officers, has extensive libraries of security equipment,
concepts of operation, site civil and landscape counter-crime features, and operational procedures-related information to help in security design planning of this
interstitial area between the protected building and the site perimeter.
The third layer of defense is the building itself, and it is here that security
design measures can play a particularly effective role. Hardening structures, facades, and systems; providing a broad array of alarm and surveillance tools;
and carefully designing and locating utilities and building systems, incorporating equipment that supports evacuation and recovery, are examples of critical
elements in security design. Subsequent chapters of this handbook expand on
topics such as the role of materials-performance criteria, structural systems design, facade protection, and survivability of ERR systems.
Task 2-2. Identifying the Critical Assets. The second task of the asset value
assessment is to identify critical elements within each of the three defense
layers—in other words, which resources must withstand an attack if the health
and safety of the building’s occupants are to be ensured. FEMA and ASIS,
among others, publish extensive checklists that will greatly assist in pinpointing relevant issues.
Task 2-3. Identifying the Building Core Functions and Infrastructure. The security design team should next identify the building’s, site’s, or venue’s core
functions and infrastructure. The design team should use these data to inform
designs that isolate, distribute, or replicate the equipment and processes that will
5
A discussion of CPTED, including numerous references and external links, can be found at
http://en.wikipedia.org/wiki/Crime prevention through environmental design.
6
ASIS can be found at www.asisonline.org.
34
DESIGN CONSIDERATIONS
provide vital operations and services during and after an attack, or designs that
use hardening strategies to improve the operational survivability of these operations and services.
One of the most successful strategies for the protection of building core functions and infrastructure in blast-protection design is dispersal and redundancy. If
a bomb detonates adjacent to or within a facility, phones and telecommunications
equipment, security and evacuation systems, fire suppression and smoke control
systems, generators and uninterruptible power supplies, and other infrastructure
elements deemed critical to evacuation, rescue, and recovery must continue to
function. Here again, FEMA, ASIS, and others have developed extensive lists of
questions to help assessment teams identify core security and life-safety functions and their interdependencies.
Task 2-4. Determining the Asset Value Ratings. The final task in Step 2 is to
assign a value rating to each asset. The value rating is the security design team’s
judgment of the degree of debilitating impact that would be produced by the
incapacitation or destruction of that asset. In other words, the consequences of
damage or destruction are considered for each asset. Then a 1-to-10, very-highto-very-low value is assigned to each asset. “Very high” would mean that the loss
or damage of the building’s assets would have exceptionally grave consequences,
such as extensive loss of life, widespread severe injuries, or total loss of primary
services, core processes, and functions. On the other hand, a “very low” ranking
indicates that the loss or damage of the building’s assets would have negligible
consequences or impact. As was previously noted, there are five additional gradients between these two extremes (Federal Emergency Management Agency
2005, Table 2.5).
The asset value assessment lists the building’s core functions and critical infrastructure elements. Next, the team assigns each element an intrinsic value
based on the degree of debilitating impact that its incapacity or destruction would
create. The product of the value asset rating exercise is a table that lists the various types of attacks against each of the facility’s critical functions. The team
then populates the table with high-to-low values expressing the consequences of
loss or damage of each asset.
2.4.3 Step 3—The Vulnerability Assessment
Vulnerability is any weakness that aggressors can exploit to cause damage. To
constructively influence design considerations, the vulnerability assessment must
realistically appraise the building’s projected functions, systems, and site features; it must pinpoint structural weaknesses and identify absences of necessary redundancy so that corrective actions and mitigations can be incorporated.
Unlike threats, vulnerabilities are conditions and/or designs that people create
themselves and over which they can therefore exercise some control. Close coordination among the assessment team, the building design professionals, and
ownership is critical; it is essential to prevent a great deal of incompatible expectations and design deliverables down the road.
COLLABORATING TO ANALYZE RISK
35
There are four tasks in assessing vulnerability:
Task 3-1. Organizing Resources to Prepare the Vulnerability Assessment.
Task 3-1 has two subparts:
(a) Selecting the Assessment Team. Risk assessment is a complex undertaking with high stakes in terms of consequences and costs. Proper team selection requires some serious up-front time and attention, if only because,
in unskilled hands, risk assessment tools can be blunt instruments. The
best survey-tool authors stress that “teams created to assess a particular
building should be senior individuals who have a breadth and depth of
experience in the areas of civil, electrical, and mechanical engineering;
architecture; site planning and security engineering; and how security and
antiterrorist considerations affect site and building design” (Federal Emergency Management Agency 2005, ii).
(b) Determining the Level of the Assessment. The security community generally recognizes three increasingly detailed levels, or tiers, of risk assessment (explained below), so at this point a decision is required: how
detailed must the evaluation be?
FEMA 452 guidelines for vulnerability assessment denote three assessment
levels:
Tier 1—A study that identifies the most important vulnerabilities and their
mitigation options—a “70 percent” assessment. A Tier 1 evaluation is typically conducted in approximately 2 days by one or two experienced assessment professionals working with the building ownership and key staff.
It involves a quick look at the site perimeter, building, core functions, infrastructure, and any related drawings and specifications. A Tier 1 assessment is likely sufficient for the majority of commercial buildings and other
noncritical facilities and infrastructure.
Tier 2—A full on-site evaluation by assessment specialists that provides a
robust evaluation of proposed and/or existing systems’ interdependencies,
vulnerabilities, and mitigation options—a “90 percent” assessment solution. A Tier 2 assessment typically requires three to five assessment specialists. It can be completed in about 5 days and requires significant key
building staff participation; access to all existing site buildings and areas,
systems, and infrastructure; and/or an in-depth review of proposed building design documents, drawings, and specifications. A Tier 2 assessment
should be sufficient for most high-risk buildings such as iconic commercial buildings, government facilities, cultural and educational institutions,
hospitals, transportation infrastructure, and other high-value targets.
Tier 3—A detailed evaluation of the facility and site. A Tier 3 analysis uses
blast and weapons-of-mass-destruction (WMD) modeling to assess the site
and building’s response, survivability, and recovery parameters. It involves
36
DESIGN CONSIDERATIONS
engineering and scientific experts, requires detailed design information,
and provides the maximum data for evaluating and developing mitigation options. Tier 3 modeling and analysis can often take several weeks
or months for an existing facility and is a process involved at the Conceptual, Schematic, and Design Development phases of new facilities. It is
typically performed only for high-value and critical structures. From 6 to
12 subject matter experts may be required, based on project complexity.
When the building and site are in the design phase and do not yet exist, the risk assessment data-gathering process described above must be
modified considerably. A Tier 2 risk assessment of a future building would
concentrate, for example, on conceptual design documents, drawings, and
plans in an iterative process of discussion, impact assessment, and design
or redesign. This situation is actually preferable, because it gives the design
professionals an opportunity to build security into the structure’s skeleton.
The alternative—retrofitting protection into inappropriate or inadequate
spaces—seldom offers totally satisfactory solutions.
The choice of the appropriate tier level depends on the building and
site’s location, its construction characteristics, its proposed use and profile,
and its owners’ or occupants’ concerns about terrorism. The choice of tier
level is also heavily influenced by the design team’s initial threat and vulnerability appraisals; by local, state, and federal antiterrorism and public
safety regulations and guidelines; and (inevitably) by benefit-versus-cost
considerations.
Task 3-2. Evaluating the Site and Building. The authors have described how
producing a robust building security design requires the team to identify
potential threats and vulnerabilities, define the resources and people most
at risk, and describe the consequences of potential losses. To bring order
and predictability out of such complexity, the team’s next task is to prepare
a survey schedule and tentative agendas. The team needs, for example,
to meet with stakeholders such as site or building owners, security and
engineering personnel, IT directors, and emergency managers. The team
will also need to inspect the area and facilities and review key information
such as security plans and emergency procedures, if these documents exist.
The key to timely execution and success is planning.
An important aid to planning is threat mapping, the clear and graphic
representation of the site-wide explosive DBTs at vulnerable points, such
as lobbies, mailrooms, and loading docks. Threat maps, which are based
on architectural concept drawings, should encompass the site from its outermost perimeter to its most sensitive internal areas. This mapping can be
done on existing blueprints or other site drawings. Each vulnerable area is
labeled with a “W” number, according to the key that has already been created for designating explosive DBTs (W-1, W-2, W-3, etc.). Threat maps
illustrate for design professionals how concentric layers of competently applied security measures can reduce, or in some cases largely eliminate, the
facility’s explosive threats.
COLLABORATING TO ANALYZE RISK
37
Task 3-3. Preparing a Vulnerability Portfolio. It just makes good sense that
the large volume of data that has been collected as part of the threat, risk,
and vulnerability assessment in the previous steps should next be gathered
into a well-organized security design document.
One of the most useful sets of assessment aids in this team portfolio
can be the FEMA Building Vulnerability Assessment Checklists found at
Appendix A of the previously mentioned FEMA publication 452, Risk Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks. This
series of tables prompts the team to adopt a consistent approach to evaluating security designs at their various stages of development. The checklist
can be used as a screening tool for preliminary design vulnerability assessments and supports the preparation of each step in FEMA’s How-To Guide.
As the FEMA Guide explains:
The checklist is organized into 13 sections: 1) site, 2) architectural, 3) structural systems, 4) building envelope, 5) utility systems, 6) mechanical systems, 7) plumbing and gas systems, 8) electrical systems, 9) fire alarm
systems, 10) communications and information technology, 11) equipment
operations and maintenance, 12) security systems, and 13) the security master plan. To conduct a vulnerability assessment or preliminary design, each
section of the checklist should be assigned to an engineer, architect, or subject matter expert who is knowledgeable and qualified to perform an assessment of the assigned area. Each assessor should consider the questions and
guidance provided to help identify vulnerabilities and document results in
the observations column. If assessing an existing building, vulnerabilities
can also be documented with photographs, if possible. The results of the 13
assessments should be integrated into a master vulnerability assessment and
provide a basis for determining vulnerability rating during the assessment
process. (Federal Emergency Management Agency 2005, Appendix A)
Task 3-4. Determining the Vulnerability Rating. The weaknesses the security
design team has identified, which could be exploited during an explosives
attack, are reflected in the vulnerability rating. As an example, the lack of
critical-systems redundancy raises the vulnerability rating, of course, as do
single points of failure among essential security and life-safety systems or
structural systems incapable of withstanding the project explosive DBTs
or an unforeseen event that could compromise a key primary structural
element.
Like the previous threat and asset value ratings, vulnerability scores are commonly expressed in a 1-to-10 numerical scale or a very-low-to-very-high linguistic scale. A very high vulnerability rating is delivered from identifying one
or more major weaknesses that make the facility extremely susceptible to an
aggressor or hazard. A very low rating describes a building that integrates excellent physical security and comprehensive redundancies—in short, a building that
38
DESIGN CONSIDERATIONS
would promise a high degree of physical and intellectual asset protection and be
operational immediately after an attack.
To summarize where the discussion is now in the risk assessment process:
In light of the identified potential threats (Step 1) and the team’s categorization
of the structure’s critical assets (Step 2), the vulnerability assessment (Step 3)
determines the opportunities to place at risk the project assets, and therefore sets
the stage for (Step 4) the risk assessment, wherein the team and the client discuss
and agree to the extent of acceptable risk.
The vulnerability assessment is a comprehensive listing of all weaknesses an
aggressor could exploit to damage an asset. The product of Step 3 is another
table, this one listing the building functions, each one with its high-to-low vulnerability rating, together with the associated vulnerabilities that the team has
uncovered.
2.4.4 Step 4—The Risk Assessment
This is the step in which the team pulls together all the information it has compiled, all the tables it has completed, all the judgments it has made, and each of
the threat, asset, and vulnerability ratings it has carefully assigned. This step is
where the interdisciplinary teamwork pays off; the risk assessment, backed up
by its demonstrable research, provides a credible foundation for security design
decisions.
Task 4-1. Preparing the Risk Assessment Matrices. Focusing attention on explosive event protection naturally narrows the range of security-design considerations. If, however, the team wants to undertake an all-risk, all-hazard review, including chemical/biological/radiological attacks as well as cyberassaults, it can
refer to FEMA 452 for a range of helpful checklists. Whether the analysis is
broad or narrow, however, the way to estimate potential losses is to prepare a series of matrices, tailored to the individual task at hand and based on the analysis
conducted during Steps 1, 2, and 3.
To estimate risk, the team must have analyzed a number of factors. First, it
will have identified and ranked the threats that could harm the project. Next,
it will have determined the value of the client’s assets, including people who
require protection. Finally, by using the results of the threat scores, asset value
estimations, and vulnerability ratings, it will determine a vulnerability and asset
rating system that identifies weaknesses an aggressor could exploit.
Task 4-2. Determining the Risk Rating. The design team next analyzes the
threat ranking, asset value, and vulnerability rating developed in the previous
steps and assigns a risk level for each critical asset. A specific risk such as a car, truck-, or person-borne bomb can be expressed as the sum of the asset value
times the threat rating times the vulnerability rating, or R = AV × TR × VR.
This exercise also results in linguistic or numeric values, or both.
Task 4-3. Prioritizing Vulnerabilities with FEMA’s Building Vulnerability Assessment Checklist. With a risk level now assigned to each critical asset, the team
is ready to prioritize, or rank, the vulnerabilities it has identified. Prioritization is
COLLABORATING TO ANALYZE RISK
39
based on evaluating which vulnerabilities the aggressors will most likely exploit
and which of them pose the greatest danger to lives, structures, and operations.
The highest ranking is given, obviously, to the greatest vulnerabilities with the
greatest consequences. Doing so makes it possible, in the next step, to determine
the most effective mitigation measures.
2.4.5 Step 5—Considering Mitigation Options
After completing the risk assessment, stakeholders are frequently left with issues
requiring mitigation, but they find their options limited by space constraints, esthetics, regulatory requirements, or costs. Decisions about allocating resources
therefore must focus on the most practical mitigation options.
These are the fundamental mitigation questions:
1. What kind of action should be taken in response to the danger? There are
four ways of responding to a threat:
r Acceptance of a threat is a rational alternative that is often chosen when
the threat has a low probability, low consequence, or both.
r Prevention is the alteration of the target or its circumstances to diminish
the risk that the terrorist or criminal will succeed.
r Interdiction is any confrontation with, or influence exerted on, an attacker to eliminate or limit its movement toward causing harm.
r Mitigation is preparation so that, in the event of an explosion, its consequences are reduced.
2. Does the response create new risks to the asset or others? The final step
in analyzing the security program’s efficacy is to be aware of new risks
created by the prevention, mitigation, or interdiction of the threats under
consideration.
Four tasks are involved in identifying mitigation alternatives:
Task 5-1. Identify Preliminary Mitigation Options. This exercise identifies the
design team’s available design options together with any relevant physical,
technical, cyber, and procedural security alternatives. The goal is to create
the strongest possible security design.
Task 5-2. Review Mitigation Options. Since every project has budget constraints as well as operations and maintenance limitations, it is seldom
possible to totally eliminate a risk. Now is the time, therefore, to weigh
mitigation options for their practicality and to match the most feasible mitigation opportunities with the most critical risks.
Task 5-3. Estimate the Cost. Some costs of protecting buildings are fixed,
while others are variable. Funds for deploying protective equipment and
creating a security zone-of-control are generally fixed; for example, the
40
DESIGN CONSIDERATIONS
costs of anti-ram barricades to keep a vehicle away from the facility are
comparable whether the same vehicle contains 50 pounds of TNT or 500.
The blast-protection costs of facility construction, on the other hand, will
vary according to the estimated threat level. Since the bomb size is outside
building planners’ control, the tailoring of blast protection to the building
structure, for instance, is a function of the maximum credible explosive
charge plus its estimated proximity when the bomb detonates.
Since cost is almost always a central issue, there are three mitigation approaches that can be used singly or in combination: (1) modify the design threat,
(2) reduce the level of protection, or (3) accept the risk. Life-cycle costs need
to be considered as well. For example, the recurring costs of maintaining and
upgrading security and life-safety equipment can derail even the best protective
programs.
Security design experience has repeatedly shown that attention paid to defending against man-made hazards during the preliminary planning and conceptual design phases significantly reduces the life-cycle costs of protection and
increases the inherent security level provided to the building occupants.
In summary, the project security professionals’ explosive threat assessment
should figure prominently in the design of the facility facade, structural system,
ERR systems, and other critical protective features. The design team should be
directly involved in the risk assessment steps that identify which building design
features and components can affect the type, size, and potential delivery method
of an explosive device.
Task 5-4. Reviewing Mitigation, Cost, and the Layers of Defense. FEMA 452
includes a final series of tables containing examples of mitigation options
and alternatives for each of the facility’s security features. These security
measures are organized by layers of defense and arrayed by degree of protection, cost, and effort. See Figure 2.3 below for FEMA 452 (January
2005) Worksheet 5-1: Preliminary Mitigation Options.
When the team has completed the risk assessment process, it will have identified and evaluated various mitigation options it can use as guidance when developing the building’s protective design features and components.
2.4.6 The Continuing Role of Risk Management
The goal of stronger, safer structures begins with building security design and
continues as security risk management throughout the building’s life cycle. To
support these initial and long-term goals, FEMA 452 asserts that the goal of the
assessment process is to achieve the level of protection sought through implementation of mitigation measures in the building design. These measures may
reduce risk by deterring, detecting, denying, or devaluing the potential threat
element prior to or during execution of an enemy attack. The Department of
Observation 1
Observation 2
Observation 3
Observation 4
Observation 5
Observation 6
Observation 7
CBR Protection and Control Measures
Blast Protection and Control Measures
CBR Repair and Strengthening of Existing
Structures
Blast Repair and Strengthering of Existing
Structures
CBR Regulatory Measures
Prioritized
Observations
Blast Regulatory Measures
COLLABORATING TO ANALYZE RISK
41
Worksheet 5-1 will help to
identify your preliminary
mitigation options.
After you have priortized your
observations/vulnerabilities
(Task 4.3), proceed to rank
them for impact during blast
and CBR events. Using the
first part of the Worksheet
(Prioritized Observations)
indicate if these observations
merit a regulatory,
rehabilitative, and/or
protective measure and if they
are directed at blast or CBR,
Using the second (Preliminary
Mitigation Options for Blast)
and third (Preliminary
Mitigation Options for CBR)
parts of the Worksheet,
determine mitigation options
that address the main
concerns included in your
observations and provided
parameters.
Preliminary Mitigation Options for Blast
Mitigation 1
Mitigation 2
Mitigation 3
Mitigation 4
Preliminary Mitigation Options for CBR
Mitigation 5
Mitigation 6
Mitigation 7
Mitigation 8
Figure 2.3 Worksheet 5-1: Preliminary Mitigation Options (From FEMA 452, Jan.
2005)
Homeland Security endorses the use of the following methodology (available in
FEMA 452).to achieve this purpose.
Deter: The process of making the target inaccessible or difficult to defeat with the
weapon or tactic selected.
Detect: The process of using intelligence sharing and security services response to
monitor and identify the threat before it penetrates the site perimeter or building
access point.
42
DESIGN CONSIDERATIONS
Deny: The process of minimizing or delaying the degree of site or building infrastructure damage or loss of life or protecting assets by designing or using infrastructure and equipment designed to withstand blast and chemical, biological, or
radiological effects.
Devalue: The process of making the site or building of little or no value or consequence, from the terrorists’ perspective, such that an attack on the facility would
not yield their desired result (FEMA 452, 2005).
2.5 CONSEQUENCE MANAGEMENT
This section provides insight into one of the primary tools used in assisting design professionals and their clients in the development of decisions involving
risk management. This section explores the value and process of evaluating consequences associated with explosive events and the major mitigating strategy
categories used to reduce those consequences.
Consequential effects are, by their very nature, subjective evaluations and categorizations. The perception of what is and what is not valuable and/or essential
is very much determined by the valuation process utilized by the institution or
organization that is responsible for its codification. Consequently, institutional
business mission statements, private organizational charters, government agency
programs, stockholder expectations, legal proscriptions, and numerous other
guidelines and/or mandates often identify different levels of expectation for postevent facility performance. These guiding documents and/or unwritten notions
were generally not created for the purpose of assisting design professionals by
clearly defining ownership’s requirements for preserving institutional businessmission operations or the protection of people on premises. The absence of wellcodified protection expectations leaves the design team with less than adequate
information about the extent of consequential effects that ownership will find
tolerable.
Design professionals must clearly understand the goals and expectations
of the facility, its occupancy and the activities that will take place within
the structure, site, and, in some cases, the campus or regional landscape
or cityscape fabric they are creating for (and with) their client. This is an
essential piece of early project information client data mining, which is often overlooked. Insufficient efforts and pursuit of a clear definition of client
desires regarding consequence management are one of the most significant
design process shortfalls. Mismanagement of this process will likely predestine the engineering efforts to fail to meet client needs or expectations.
To address this omission in the programming design phases, one of the first
steps should be—during the earliest phases of project design programming and
conceptualization—the creation of a brief security design mission statement.
This document should clearly and concisely state the client’s expectations for
the facility’s post-explosive event performance. These expectations are routinely
overstated by clients. This is generally a result of their lack of sophistication or
CONSEQUENCE MANAGEMENT
43
knowledge of the engineering complexities, architectural program compromises,
post-construction institutional operational constraints, impact on project financials, aesthetic effects, and other factors better understood by the structural engineer and the balance of design team professionals. Consequently, this mission
statement must be a consensus-developed document. It is reasonable to expect
that it may take several iterations of authorship, editing, review, and subsequent
approval.
Once agreed to, the consequence management statement should then become a part of the project’s key design assumptions and design program
objectives. It should be used as a means to guide design decisions, and to
provide the basis for decisions regarding program conformance, as designs
mature in the Schematic, Design Development, and Construction Document
phases.
Once a clear concept of acceptable consequence management in terms of facility performance is understood and codified, there is a sound basis for developing the facility designs. This is essential to the process of achieving an acceptable
balance between post-explosive event damage to the facility’s physical and intellectual assets and the expectations of the client and society regarding life safety
and institutional business-mission continuity.
No facility can be expected to provide a post-explosive event damage performance profile that ensures no injury, no loss of life, and no disruption to
facility operations. The challenge of consequence management is for the
design professional and client team to establish reasonable goals and objectives. These need to be tempered by rational opportunities for success
given the facility’s design program, client occupancy and activities, project
location and budget, law enforcement and security resources, and other economic and geopolitical factors, which are unique and specific to each design
commission.
Design professionals and their clients have opportunities to examine and implement strategies that can participate in achieving the level of consequence management identified in the project’s key design assumptions and design program
objectives, as stated above. These holistic design strategies usually include:
(a) Strategic site selection and facility location to reduce vulnerability to
threats that are on-site, near-site, or regional.
(b) Within a specific building, the provision of equipment, personnel, and resource redundancies, so that the loss of a single critical system, facility
management operational component, or structural element or system does
not create a single point of unacceptable or catastrophic failure.
(c) Within a specific building, distribution of facility physical plant; intellectual and nonintellectual assets; operational programs; and evacuation,
rescue, and recovery systems so that an explosive event in one location does not jeopardize the entire building, all of its occupants, and the
facility-wide performance of ERR systems. (Note, however, that the use of
44
DESIGN CONSIDERATIONS
diverse redundancy as a protective strategy should be examined during the
design process in conjunction with the protective strategy of co-location of
critical systems and/or resources housed within a well-protected environment. The intent is to foster survivability, thereby reducing consequential
effects.)
2.5.1 Consequence Evaluation
It is universally understood by the architectural and engineering design community that explosive events can be responsible for significant damage to building
structures, facades, and critical building systems. However, there is less awareness amongst the design profession during the initial planning phases of other
extremely important and sometimes ethereal values such as an institution’s public image and sense of solidarity. Without this awareness there is naturally less
acceptance of design responsibility for the protection of these other, less tangible
but extremely important client assets.
Additionally, other more essential issues associated with evacuation, rescue,
and recovery operations generally are assumed by the design profession to be
owner and operator responsibilities developed by those entities in response to
the building’s design. This approach is archaic.
Design professionals must address building designs so that the explosive
event consequence management response is achieved through an evaluation
of how evacuation, rescue, and recovery operations will occur after the explosive event. In this way, designs are informed to protect evacuation, rescue, and recovery building design features such as stairways, exit passageways, elevators, and building systems such as emergency power, emergency
lighting, voice evacuation, fire detection, fire suppression, and smoke management. This consequence evaluation approach results in a building design
that provides inherent tactical support and definitive criteria for an evacuation, rescue, and recovery concept of operations. This will always provide
a superior product to an ERR strategy retrofitted to a design created with
less inherent vulnerability.
The above design shortfalls can generally be attributed to the design team’s
focus on the intimate knowledge of and extensive involvement in developing
all of the intricate details associated with each component of the building’s architectural, structural, and support system infrastructure designs. Furthermore,
the design community traditionally focuses on contract boundary limits as the
extent of their responsibility for the ultimate scope of their work. The range
of effects from explosions rarely aligns with contract boundary limits of
work. This fact must be taken into consideration as part of consequence
evaluations.
Facility occupancy after construction completion, with its complement of staff
population assignments, support furnishings and equipment, and operational protocols usually is established with the greatest attention to detail after the building
is commissioned. Consequently, these and other aspects of day-to-day facility
CONSEQUENCE MANAGEMENT
45
management are logistically more distant entities and often are understood in
less detail by the design professional team. Design professionals should focus
additional resources on more completely understanding post-construction
occupancy and fit-out subtleties, in order for explosive event consequence
evaluations to be more thoroughly responsive.
Also, traditional contracts, governing the performance of the design team’s
scope of work, emphasize the most significant levels of effort up to and inclusive
of building commissioning. They fall short of emphasizing the role of the design
professionals in fully understanding a broader range of construction completion
and subsequent facility occupancy concerns. These include all of the various intellectual and operational entities, exceptional conditions and crisis management,
and facility restoration issues, which will be placed at risk and/or hindered by an
explosive event once the structure is completed. Understanding how the facility will operate under extraordinary conditions is essential to developing
protective designs that address the consequential effects on occupants and
operations, as well as on the physical plant.
The structural engineering, architectural design, and balance of the design team professionals should ensure, at the commencement of the design
process, that this disparate intellectual appropriation of design awareness is
addressed. This should be accomplished through a process that focuses on
the protection and post-explosive event performance of the structure itself
in a manner that is carefully balanced against all of the other consequential explosive event effects, including day-to-day facility operations and the
exceptional conditions associated with ERR.
A wide range of consequences must be evaluated. Understanding these consequences is critical to establishing the balance between (a) protective designs
assigned to the facility’s site perimeter protection program, facade, structure,
evacuation, and rescue and recovery systems and (b) systems associated with
life safety and business continuity. Mitigating each of the consequences that
may occur if these areas of facility vulnerability are not adequately addressed
will require a portion of the project’s budget. The project profile for security
investments will in large part be determined by the consequence evaluation
process. This process can be logically approached and informed by an evaluation of the following areas of explosive event influence:
Site and Local Risk Management Context Although design professionals
are traditionally charged with protective designs for the specific contracted-for
client facility, explosive events routinely involve collateral damage to adjacent
structures and/or public spaces and also place these additional lives and property
at risk. This is a fact that cannot be ignored. The performance of the client
facility must be considered within a regional context in order for the client and
the design team to exercise the highest standard of industry care and to reduce
the extent of design professional liability. The creation of a highly protected
facility that exclusively focuses on protection of that specific facility’s assets
and does not take into consideration how its performance may affect adjacent
46
DESIGN CONSIDERATIONS
structures, public spaces, and their populations cannot be considered the most
prudent design solution.
Facility Structural System Protection of the facility’s primary and secondary
structure from progressive collapse and disproportionate damage associated with
explosive events external or internal to the building is a baseline protective designconsideration in consequence evaluations. Catastrophic failures of an entire
structure or a significant portion of it generally place large occupant populations
at risk, both as a result of the immediate event effects and of disruption to ERR
operations. Additionally, these structural failures invariably cripple critical business missions, sacrifice institutional operational viability, and adversely affect
restoration of day-to-day operations.
Facility Facade System Explosive events, either internal to or external to the
building, are likely to engage the facade in managing applied forces significantly
in excess of the thresholds associated with conventional wind, seismic, and gravity loading. This routinely results in catastrophic glazing and glazing system
failures, which are historically responsible for significant loss of life and injury
to both occupants within and external to the targeted facility. Airborne glass
fragmentation continues to be identified as one of the most lethal aspects of
building component responses to explosive events.
Facility Evacuation, Rescue, and Recovery Systems While the immediate devastating effects of explosive events can be limited by protective design strategies,
post-event evacuation functions and the subsequent activities of rescue and recovery are dependent upon the integrity of stairs, vertical transportation systems,
power, lighting, voice communication, smoke management, and other systems.
These systems can be placed out of service throughout the building or suffer
disproportionate damage if they are not specifically designed in response to the
project design basis explosive threats. Maintenance of these critical ERR systems, outside of the immediate explosive event zone, is essential to reducing
the extent of resulting lethality, injury, and facility disruption. Design professionals, when evaluating explosive event consequences to ERR systems and
operations, must take into consideration that these systems may be exposed to
supplemental damage due to structural instability. Also, that extraordinary (terrorist) criminal event modalities have historically been strategically planned so
that phased explosive events occur with the purpose of targeting emergency responders and those residual occupants on the scene who have not yet been evacuated.
HVAC Systems Explosive events are traditionally accompanied by the development of significant quantities of airborne particulate contaminants, including
glazing fragmentation, toxic materials, and other byproducts of building component demolition. These are inevitably deposited within the HVAC system as a
result of system breaches associated with the explosive event, or as a result of
CONSEQUENCE MANAGEMENT
47
fans continuing to operate after the explosive event. This contamination can be
of such significance as to warrant the demolition and entire replacement of the
HVAC fan systems (including filters, coils, etc.), fan plenums, and distribution
duct work.
Business Continuity The design professional community should naturally address issues of life safety as a first priority. However, considerations for institutional business continuity cannot be ignored as part of the initial consequence
evaluation process, and may in fact be a client’s most clearly stated objective.
The extent of the business continuity protective design mandate will depend upon
the client’s mission statement and expectations. It will also be dependent upon
the extent of resources available to provide building protective design strategies beyond those required to achieve the consensus of life-safety performance
standards. In general, clients’ expectations regarding the financial benefits
of business continuity are well stated. However, ownership routinely anticipates that the building codes and design professionals’ experiences will adequately guide the design in the protection of occupants on site. This potential
disparity in consequence management expectations must be identified and
addressed accordingly.
Restoration of Operations Explosive events are violent conflagrations with the
potential to significantly disrupt facility physical plant, affect life safety, and
violate viable occupancy. Every explosion will be initially treated by the law
enforcement community as a crime scene, by the building department as a potential condemnation, by underwriters as a time for claim investigations, and
by others as the site of numerous activities other than those originally intended
as part of the building’s use profile. There will be an inevitable period of investigation and severe facility occupant access restriction. Design professionals
should engage the client in post-explosive event consequence evaluations.
These should very specifically focus on the extent of post-event facility demolition and restoration that is tolerable in order to meet client expectations regarding facility condemnation by the local building authority, the
anticipated time and cost of reparations, and the extent of time required to
reestablish baseline business-mission operations.
Target Attractiveness To the extent that crimes involving the use of explosives
can be deterred, a significant advantage can be achieved. While there are no specific predictive mechanisms that can be referenced as an engineering and design
guideline for criminal deterrence, especially for extraordinary (terrorist) attacks,
the concepts of target hardening and inaccessibility are universally accepted by
the law enforcement and intelligence agencies as rational approach strategies for
crime reduction. Consequence evaluations should consider, at all stages of
design, whether the facility will be an appealing target. It is especially important that target attractiveness be considered in the earliest phases of site
selection and facility location within the site. Also, the extent of accessibility
48
DESIGN CONSIDERATIONS
and opportunity for an aggressor to deliver the explosive to the target needs
evaluation. Reduction in the perceived opportunity to achieve both of these potentials is a design goal worthy of substantial consideration.
2.5.2 Function Redundancy
This subsection identifies and provides insight into the benefits associated with
the management of post-explosive event consequences through the application
of redundancy in various building architectural elements, programmatic spaces,
and/or building systems. Incumbent on the discussions regarding redundancy is
the necessity to advance the concept of locational diversity, as adjacent redundancies are not efficient managers of consequences from an explosive event. Simply
stated, adjacent assets are most likely to suffer the same damages.
This subsection text includes the identification of building systems involved
in evacuation, rescue, and recovery, and discusses how redundant ERR systems
may reduce the requirement for structural hardening or may otherwise inform
structural concepts, while reducing consequential effects. This is essential as
ERR systems have a primary role in enhancing facility life safety in response
to an explosive event and the secondary consequences of such a circumstance.
This subsection also addresses business mission-critical systems and spaces
and how redundancy and diversity of these systems may beneficially affect structural and facade designs from the standpoint of explosive event protection. Other
aspects of redundancy associated with building components, systems, or program spaces are also discussed with a focus on how they may inform the structural designs in terms of acceptable post-explosive event effects.
The subject of structural redundancy, for primary and secondary structural
elements, is not covered in this subsection, as it is the specific subject matter of
other chapters in this handbook.
A primary facility performance goal for the design professional is to ensure
that evacuation, rescue, and recovery systems remain operational outside of the
specific area of influence associated with the project-specific explosive design
basis threat event. This is a significant challenge since there are a number of
systems that participate in ERR. Design professionals should anticipate that the
following systems, at a minimum, are considered part of the ERR protection
strategy:
1.
2.
3.
4.
5.
6.
7.
Fire suppression systems, including their water supplies
Fire alarm, detection, exiting, and annunciation systems
Emergency voice communication systems
Emergency lighting systems
Video surveillance systems supporting ERR spaces and operations
Emergency way-finding and exiting systems (other than exit signage)
HVAC chemical, biological, and radiological (CBR) agent control
CONSEQUENCE MANAGEMENT
8.
9.
10.
11.
12.
13.
49
HVAC smoke control, purge, and pressurization systems
Elevators designated for Fire Department use
Access control systems engaged in emergency exiting
Fire and security command and control facilities
Emergency power systems supporting the systems listed above
Emergency exiting stairs and passageways
These are complex systems, with components distributed throughout a building. The opportunity for their disruption is high as a result of the high-energy
devastation characteristic of an explosive event. Consequently, their survival outside of the immediate explosive event zone will depend upon significant investments in hardening of walls and slabs to protect both the front-end monitor and
control portion of these systems and their extensive and pervasive distribution infrastructure. Consequently, this is rarely the most economical approach. It is also
least likely to provide reliability equivalent to that of a design that implements
redundant systems located with geographic diversity, whose extent of segregation is based upon informed explosive event calculations and modeling. In summary, placing critical ERR system components outside of the blast effect
zone is more often a more survivable and robust solution than attempting to
provide a hardening strategy to overcome single points of vulnerability and
consequent failures.
Business-mission continuity depends upon the survival of critical systems,
equipment, information, and intellectual resources. Consequently, they must be
located within safe and secure spaces that are appropriately designed for human
occupancy and provide conditions that do not exceed the systems’ thresholds of
environmental tolerance. These are fragile components, systems, and/or conditions that are as susceptible to disruption or compromise from an explosive event
as the previously discussed ERR systems. Depending upon the type and extent
of service and/or product provided by the institution, there may be a requirement for protecting a particular portion of or, alternatively, the entire facility.
Design professionals, in the Preliminary Planning and Conceptual Design
phases, will need to partner with the client and determine the extent of the
facility that must remain free of explosive event effects (as well as exterior
support infrastructure) in order to achieve the client business continuity
expectations.
If a backup project site protective design approach is not identified as a viable
option (see Section 2.5.4), then the strategy of individual physical and intellectual asset protection will need to be compared to a strategy employing redundant
systems and personnel support within the project facility. These will need to be
located at a dimensional separation established by informed explosive event calculations based upon the project explosive design basis threats and the locations
where these threats can be delivered. These locations will be dependent upon the
facility’s security program. Where local hardening options exist, they should be
50
DESIGN CONSIDERATIONS
carefully examined and a determination made as to whether they are the most
cost-effective and reliable approach to a robust business continuity design concept. If so, then critical business operations and their support systems can be
confined to a compact and unified occupancy program within a locally hardened
portion of the structure.
External explosive events may place significant portions of a building’s facade at risk, and the failure mode may involve large portions of the exterior
envelope. Based on the explosive event specifics, including the threat assessment
explosive DBTs, standoff, and architectural envelope design criteria, post-event
facade response may involve significant inbound facade debris that could negatively impact life safety, ERR operations, and business continuity programs.
Design professionals, in the Preliminary Planning and Conceptual Design
phases, should partner with the architectural team and develop preliminary
ERR design strategies, informed by facade explosive event experience, so
that determinations regarding facade hardening versus redundant diverse
ERR systems are given due consideration and the proper balanced solution
selected.
While ERR components and operations and business continuity mission expectations traditionally receive the majority of design focus for protection from
an explosive event, client requirements may dictate that other physical, intellectual, or operational assets survive the DBTs. To address this additional requirement, design professionals will need to clearly understand the type and locations of assets at risk. Subsequently, the post-explosive event architectural,
structural, mechanical, electrical, and other related systems’ performance requirements will need to be defined in a manner responsive to the asset protection
profile. Based on the complexity of the necessary support infrastructure, design professionals should conceptually investigate the option of individual
system hardening versus redundant support systems at geographically diverse locations outside of the calculated explosive event effect zone. Even if
a redundant diverse support system is selected, it should be recognized that local hardening will likely be required at points of connectivity to the protected
space(s).
In some cases, functional redundancy is not an option. The preservation of
unique cultural artifacts, the protection of unique financial assets, co-located intellectual resources, critical unique industrial or financial processes, and other
single-source assets do not allow the design professional the benefit of protection through diverse repetition. These will require competent hardening strategies specifically created in a manner responsive to each asset’s vulnerability.
A discussion of functional redundancy and diversity, and mitigation strategies involving an analysis of hardening versus repetition and separation is not
complete without consideration of effective strategies to preclude the introduction of explosives to the facility or to assets at risk, through the implementation of a competent security program. Design professionals should vigorously
pursue with ownership opportunities for threat reduction. The potential for
implementing this approach will yield a more secure facility with greater
CONSEQUENCE MANAGEMENT
51
functional capability for ERR and the least threat to business continuity or
other single-source, high-value assets.
2.5.3 Building Location
This section explores and defines the opportunities that the structural and blast
consultant engineers have available to assist the client in understanding the structural, architectural, site civil, constructability, cost, schedule, and operational impacts associated with explosive event protection during the site selection process.
Design professionals are generally given the opportunity to assist a client in
site selection, or, alternatively, are asked to develop the best structural protective
design response for a building on a particular client preselected site.
Site selection criteria will be influenced by the threat assessment as identified
in Section 2.4 of this handbook.
When design professionals are engaged to work with the client on identifying the suitability of a site, or sites, and the placement of a structure on
those site(s), the decision should be informed by a competent and thorough
threat and risk assessment. This assessment must clearly address the qualitative
explosive threats (vehicle-borne, man-portable/satchel, maritime vessel– delivered, airborne delivery, standoff military ordnance, etc.), the quantitative design
basis threat values (amount of TNT equivalent for each of the threats identified,
shaped or unshaped charge, etc.), and the extent of opportunity for establishing
standoff between the threats and the protected facility. These are discussed in
more detail in Section 2.4.
Similarly, when clients request that design professionals work within a given
predetermined site (i.e., when informing the site selection process is not part of
the design professionals’ responsibilities), the limitations which that specific site
will impose upon the structural engineers and other design professionals should
be identified. This should be based upon the risk assessment results associated
with an assessment performed for that specific site.
Site selection and building placement locational consequences can be addressed through a competent process of explosive event analysis. This analysis
must be informed by the threat assessment. It should also be informed by standoff distance determinations, the evaluation of the resulting extraordinary loads
imposed upon the facility by the explosive event threat, and the opportunities for
mitigating strategy protective design concepts available to the design team.
Elimination of the explosive event threat remains the most attractive,
competent, and cost-effective process in developing designs for facilities on
any site. However, based on the threat assessment, it may be unreasonable to
assume that the facility will not be exposed to explosive event effects. Maintenance of standoff between the explosive device and the asset remains as one
of the most powerful mitigating strategies available to the design team for
the development of a structure that affords the best opportunities for risk
and consequence management. In the case of standoff military weaponry (such
as shoulder-fired ordnance), should it be identified in the threat assessment as a
52
DESIGN CONSIDERATIONS
design basis threat, standoff distance will likely be less important than obscuring line of fire, the provision of a predetonation screen, or the provision of other
measures, as this form of attack can take place from a considerable distance.
In general, applied pressures and the duration of their application (impulse) to
the structure drop off exponentially as the distance between the protected structure and the explosive event increases. The protection of the facility’s facade;
primary and secondary structure; evacuation, rescue, and recovery systems; business continuity–related intellectual assets and physical plant; and other preidentified entities worthy of protection will require a high level of design sophistication. The design process will need to balance budget allocations and architectural
and structural design responses based upon the achievable standoff between the
assets to be protected and the explosive event delivery location.
As part of the site selection process, the blast consultants, working with
the threat assessment identified DBTs, should work closely with the client
and design professional team. This process should establish the pressures
and impulses that will be applied to the structure, its facade, inbound utilities, and other critical site civil features, based on the maximum achievable
and realistic standoffs. This is essential so that an initial determination can be
made on how these extraordinary loads (in excess of gravity, wind, and seismic) will affect the protective design capabilities of the following major building
systems and/or afford protection to the following building components and occupants:
1. Major/critical inbound utilities, such as fire protection water service, utility
electrical power, information technology and data services, domestic water,
steam, and gas
2. Facility access control points for vehicular entry
3. Site-located critical utilities such as transformers, generators, cooling towers, fuel storage, domestic and fire service water storage, and communications infrastructure
4. Facility facade
5. Facility primary and secondary structural systems and elements
6. Vertical transportation systems, including stairs, escalators, and elevators
internal to the building (these may be exposed to explosive event effects
should the robustness of the facade design be less than the DBT event
calculated pressures and impulse loads)
7. Evacuation, rescue, and recovery systems as identified previously, in Section 2.5.2, internal and external to the building (these may be exposed to
explosive event effects should the robustness of the facade design be less
than the DBT event calculated pressures and impulse loads)
8. Occupants within the facility
9. Business mission-critical systems and equipment located within and/or external to the facility
CONSEQUENCE MANAGEMENT
53
The opportunity to achieve maintenance of standoff between the explosive
DBTs and the project assets, as noted above and as supplemented for each specific project, should be one of the primary focuses of the site selection process.
This should involve a careful evaluation of how the explosive device delivery
threat, as identified in the threat assessment, can be competently separated from
the asset(s). Rational solutions involving vehicle anti-ram perimeters, pedestrian
interdiction barriers, and other site civil and landscape features should be considered for their implementation feasibility. Inability to institute these protective
features may have consequences on the building’s design. Consequently, the following strategies should be considered:
1. To increase survivability of critical inbound utilities, and, therefore, the reliability of ERR systems and operations, business continuity, and the mitigation of secondary event effects (such as the ignition of a gas main as a
result of the explosive event), these utilities should be located outside of
an explosive event delivery zone. Alternatively, they should be provided
with hardening protection and redundant diverse routing at dimensional
separations identified by the blast engineer. Utility exposure and explosive event survivability evaluation is an initial critical design determination, as it may inform substantial work and expenditure for utility
hardening protection or diverse routing and/or establishing long lead
time approvals and/or feasibility studies by local utilities.
2. To increase protection for building occupants, ERR systems, critical
business-mission operations, and other predetermined internal assets, the
facade system for the building should be reviewed for its ability to be designed to reduce the threat of airborne fragmentation to both internal building occupants and those external to the building. This should include a determination as to whether a design can be implemented whose post-event
performance will be adequately robust to preclude additional protective design measures internal to the building, or whether the failure mode requires
supplemental protection of internal and external assets. Facade robustness
evaluation is a critical initial design determination, as it will influence
the building’s architectural aesthetic, internal space program assignments, and the dedication of additional expenditures for hardening
and/or diverse redundancy of ERR and business mission-critical systems.
3. Initial structural design concepts should be evaluated for their ability to
increase the survivability of the primary and secondary structure, and, consequently provide significant enhancements in life safety and businessmission continuity, and reduction in both decontamination and restoration
time to resume normal operations. These conceptual structural systems
should be reviewed for their ability to be resistant to progressive collapse
and disproportionate damage, based on the building location and the available standoffs. Structural system survivability is a critical initial design
54
DESIGN CONSIDERATIONS
determination, as the type of structural system for the building may,
in fact, be determined by explosive event management, in lieu of being
conceived to address traditional gravity, seismic, and wind loadings,
in order to achieve the post-explosive event performance results. If a
conventional structural system cannot satisfy the security performance requirements, then site selection and/or building placement on the site may
have blast protection requirements that inform member sizing, connections, standoff, and other major portions of the design with consequential
effects on construction schedule, budget, and the architectural program.
4. To ensure the most competent and effective means to facilitate evacuation, rescue, and recovery operations, the designers should evaluate egress,
emergency responder vehicular access, and emergency responder entry.
This should be performed through a process of establishing their exposure to the project explosive DBTs and the separations/standoffs that can
be provided by the particular site. Vehicular access points to the site, approach configurations to the facility, and the conceptual location of exit
stairs and elevators should all be reviewed for survivability. The survivability of ERR facility components is a critical design determination,
as it may inform major aspects of the site civil and landscape plan, and
the location, quantity, and type of the facility’s vertical transportation
systems, as well as the hardening required for these systems.
2.5.4 Building Dispersal/Distribution of Functional Programs
This subsection addresses the opportunities for consequence evaluation and management associated with the physical separation of internal building program
spaces, critical ERR and business continuity support systems, and physical assets and occupants. It focuses on the relationships among facility vulnerability,
facility diversity (geographically diverse facility entities and, also, redundancy
and diversity of internal building assets at a single site), the blast analysis process and the associated analytic results, and how these factors and engineering
analytics can inform whether physical separation and the resulting diversity are
a more rational protective design solution than co-location and structural, architectural, and ERR system hardening.
The threat assessment process discussed earlier identifies credible threats appropriate for mitigation. If explosive event threats are subsequently identified,
then both the site selection and facility design process will need to address their
vulnerability to the specific explosive event types, design basis threat quantification, the means of delivery, and their adjacency to the protected asset.
Vulnerabilities will occur wherever the explosive event threat can be delivered at distances determined by the blast consultant to impose pressures, impulses, soil movements, and other explosive event effects in excess of the loads
prescribed for management by the building code. Additionally, vulnerabilities
may exist where these explosive events create conditions in excess of the criteria
CONSEQUENCE MANAGEMENT
55
stipulated in the client’s published design standards for project performance. It
is essential for a qualified blast consultant to inform the design team of these
extraordinary loads, vibrations, soil displacements, and other consequential
effects, and the distances at which they will be applied to the facility’s structure, facade, and critical systems. These values should not simply be obtained
from resource and reference materials unless they are reviewed and subsequently
modified and/or approved by a qualified explosives and structures subject matter
expert with extensive experience and access to government-approved computer
modeling software.
Subsequent to the identification of the anticipated explosive event effects,
based on the type and delivery location opportunities for the explosive threat,
conceptual building placements on the site and conceptual identification of ERR
and business-mission or other critical assets should be identified by the design
professionals working with the client team. Various options are likely to exist
in the Conceptual and Schematic design phases, and each of these should be reviewed to identify explosive event vulnerability exposures. Based on the agreedto, post-explosive-event facility design survivability performance criteria, the design team should consider:
1. The development of maximized enforceable standoffs.
2. The utilization of protective design facade, structure (primary and secondary), and internal partition hardening.
3. The addition of redundant ERR and other systems located diversely enough
within the facility to ensure survivability to a single or multiple explosive
events, as defined in the threat assessment. Note that this must take into
account both external and internal explosive threat modalities.
4. The repetition of those portions of the facility at a diverse site (essentially
another standalone clone of the assets at risk), at a distance ensured to be
free from the explosive event effects at the project site as determined by
the blast consultant.
Life-safety concerns for facility occupants and those external to the building are not specifically addressed by the concept of redundancy and/or diversity,
other than through the benefit they receive from the protection afforded by survivable evacuation, rescue, and recovery systems. Those within the immediate
blast effect zone (an area of influence that generally can be well estimated by
the blast consultant) will suffer traumatically unless they are also protected by
blast-hardened elements and free from exposure to airborne debris. However,
the protection of those systems that aid ERR operations will significantly enhance the life safety of those who will need to exit the facility and provide rapid
and efficient access by emergency responders immediately after the explosive
event. These activities are essential in assisting those facility occupants affected
by the event or those building occupants who fall within the ADA-defined profiles and require supplemental assistance for evacuation. In summary, diversity
56
DESIGN CONSIDERATIONS
and redundancy are strategies best employed for the protection of ERR systems
and building features, critical business-mission systems, equipment, and operations or other client-defined physical assets. The initial phases of consequence
evaluation and management should occur in the Conceptual and Schematic
design phases and include a review of the opportunities for redundancy and
diversity and the use of hardened facility design elements.
The above notwithstanding, certain facility operations and client requirements
may dictate the duplication of staff either within a facility or at a remote site.
This will occur when individuals are considered to be of such importance that
they are an integral component of business-mission continuity support. When
this is a design parameter, the structural engineer, architect, and blast consultant must address the project’s explosive design basis threats in a manner
that identifies the extent of segregation required between the area where those
threats may occur and the proposed location for this supplemental protected
population.
This process must evolve from the development of initial “threat maps,” which
are based upon the available architectural concept drawings. These maps should
be created to identify the location of explosive event delivery opportunities, both
external and internal to the facility. This will require the parallel conceptual development of a facility security operations program. Consequently, a security
professional should be part of the design team’s composition so that security
systems and management operations are designed to limit the potential size,
delivery strategy, and locational placement of explosives.
Once there are reasonable key design assumptions regarding explosive event
threat delivery locations, explosive event analytics should be performed to determine whether a co-located population of duplicate staff (and the infrastructure
required to support them) can be adequately protected within a single building,
or whether they need to be physically relocated to a remote site. Although this
may appear to be a daunting and challenging process during the early stages of
the project, where designs are truly conceptual, this will provide the best opportunity for informing the major structural and architectural aspects of the design. The longer the process of determining how staff will be protected is
delayed, the more difficult the protective design solutions will become. This
will therefore generate a compromised set of strategies that will cost more
and provide less competent protection.
2.5.5 Disaster Recovery and Contingency Planning
Disaster recovery and contingency planning provide carefully conceived and
tested alternatives for taking action when things go wrong. Since no privatesector structure can be built to survive every explosive attack, and since no security plan is perfect, every organization that believes there is a threat to its facility or proximate environs should create disaster recovery strategies, contingency
procedures, and business recovery plans. Indeed, most boards of directors and
insurance companies demand no less.
THREAT REDUCTION
57
The purpose of contingency planning is to protect building occupants and
speed business resumption, including regaining access to data, communications,
workspace, building support systems, and other business processes. In short, the
consequence management and the other security design concepts discussed in
this handbook can and should support business continuity planning. Early in the
conceptual design process, therefore, the design team should discuss post-blast
recovery with ownership in the specific context of their contingency planning
expectations. Design professionals should insist on reviewing with the client
the opportunity to incorporate contingency planning and business recovery
concepts into their building design.
From the occupants’ perspective, the single most significant reason why contingency planning is important is that lives could hang in the balance. To the
eventual building client institution, effective business recovery strategies are also
critical in maintaining the public’s confidence in the enterprise during a crisis.
2.6 THREAT REDUCTION
The design professional, security, and law enforcement communities acknowledge that facilities can be safer environments when those threats that place the
institution and its occupants at risk are managed by a process of threat reduction and/or, preferably, elimination. When the law enforcement and intelligence
agencies cannot ensure management of exposures to an explosive event, then institutions can achieve supplemental threat reduction by being placed as far from
the threat as possible, so that the residual effects of the threat at the maximum
achievable standoff distances are then addressed by the facility’s site selection,
facility placement within the site, and the facility’s design. These relationships
will be discussed in this section.
Explosive events can be categorized as those that occur:
1. As a result of the purposeful, beneficial, and legitimate use of explosives
(e.g., rock excavation for foundations)
2. As a result of purposeful, willful, and malicious criminal acts, using explosive materials (e.g., a terrorist bombing)
3. As a result of a criminal act that does not initially use an explosive material,
but that subsequently engages other resident agents possessing explosive
capabilities (e.g., an arson event that then engages a compressed gas vessel)
4. As a result of an accidentally occurring event (e.g., a high-pressure steam
line rupture)
5. As a result of a naturally occurring event (e.g., a lightning strike engaging
an electrical transformer)
The law enforcement and intelligence communities and the security professionals responsible for securing a city, neighborhood, and/or facility site face an
58
DESIGN CONSIDERATIONS
impossible challenge when tasked with the complete elimination of the explosive
event threat. The prevalence of the suicide bomber creates exceptional opportunities for explosive event delivery that preclude providing the design professionals
with a design basis threat of “no explosives.” Consequently, when design professionals are tasked with the development of protective design strategies
for a site, building, and its occupants, they must insist on obtaining the specific explosive design basis threats, which are identified within a competent
threat and risk assessment created with contemporary input from the law
enforcement and intelligence agencies on explosive material types, delivery
modalities, and any other related available specifics. This information mining
is essential if credible structural and architectural facility designs are to provide
real-world protection.
Furthermore, it should be recognized that the law enforcement and intelligence communities cannot effectively identify the extent of explosive event
threat reduction that a facility security program will accomplish, unless, perhaps, it is a government facility for which they have direct management control
and involvement. In most cases, on-site and internal-to-building explosive event
threat reduction must be identified by the client and their security professionals, or a highly qualified security consultant. This information must also inform
the project threat and risk assessment so that quantifiable explosive design basis threats are represented on the previously discussed facility threat maps. This
process will allow a clear and graphic representation of how explosive threat
reductions can occur, starting at the facility site perimeter and migrating to the
most sensitive internal2 areas requiring protection. The intent of threat mapping is to clearly identify explosive quantity and locational placements to
inform the design professionals on how the explosive event DBTs are reduced and/or eliminated as a result of being subjected to concentric layers of competently applied security measures. This is the information which,
as represented on the threat maps, will inform how threat reduction can be
used by the design team to address malicious criminal acts involving the use of
explosives.
To reduce the threat associated with naturally occurring events, the designers should gather relevant, project specific information and review the
site and the facility designs at the earliest possible stages of design. The purpose of these investigations is to identify those sources of explosive event opportunity worthy of supplemental consideration and protection. This should involve
the identification of other explosive event substances that contain inherent highenergy sources such as:
1.
2.
3.
4.
5.
Compressed gas vessels and/or piping
Liquefied natural gas vessels and/or piping
Fuel storage and piping other than gaseous agents
High-pressure steam vessels and/or piping
Stored munitions and materials used in their fabrication
THREAT REDUCTION
6.
7.
8.
9.
10.
59
Electrical transformers
Combustible airborne particulates
Combustible chemicals and agents other than explosives
Chillers, boilers, and furnaces
Numerous others
These may occur either adjacent to or resident within the site and/or facility,
or exposures may occur as a result of person and/or vehicle (motor vehicle, rail
transit, aircraft, or maritime vessel) delivery within or adjacent to the facility.
2.6.1 Accidental Explosions
This subsection identifies potential accidental explosions that should be taken
into consideration by the design professional team. The intent is to ensure that
design professionals—especially the structural, mechanical, and electrical engineers and the architect—are aware of and provide protection for explosive events
other than purposeful, malevolent criminal activity.
Explosive events occur as a result of the release of high-energy sources. While
traditionally this condition is conceived of as the result of ignition of an accelerant, such as highly flammable liquids, compressed gases, or explosives, other
high-energy sources—such as steam or water under extreme pressure, chillers,
boilers, and other systems routinely found in facilities—are also candidates for
imparting extraordinary pressures over extremely short durations.
The unabridged edition of the Random House Dictionary of the English Language defines an explosive event as “the act or an instance of exploding; a violent
expansion or bursting. . .” and “a sudden, rapid, or great increase.” While the
pressures, impulses, and plasma effects associated with an explosive event
using extremely high-energy substances such as C4, Semtex, RDX, TNT, or
their equivalents, are regarded as upper-level design thresholds for protection, design professionals should carefully review the entire project design
with the intention of identifying other building systems and/or programmatic spaces allocated to the storage and/or distribution of substances and
materials that could participate in an explosive event. These systems and areas of programmatic use are unlikely to be identified by a traditional security
threat and risk assessment, as these activities usually focus upon the purposeful
introduction of high-energy explosives as the anticipated design basis threat.
At the earliest possible design phase, the structural engineer, mechanical and
electrical engineers, and architect should review the facility design and identify
whether any of the systems, equipment, or environmental conditions identified
previously are resident within the project scope. This process may involve making specific inquiries of the client, based on the sophistication of their occupancy
program, regarding what explosive exposures they know exist, as a result of their
detailed familiarity with the facility operations that will occur in the project. The
design team should list, as part of the protective design strategy’s key design
60
DESIGN CONSIDERATIONS
assumptions, the potentially explosive operations, systems, and/or materials
and substances that will be present as part of the building’s use.
A list of potential explosive event conditions should be created based on the
aforementioned plan review and client inquiry. Individual mitigating strategies
should be proposed by the design team to address these exposures. This may
include the following:
1. Elimination and/or reduction of explosive sources and/or quantities from
the project
2. The introduction of detection systems to identify the conditions associated
with the development of an explosive event condition prior to its detonation
and/or release point
3. The introduction of suppression systems capable of reducing the postexplosive-event effects
4. The introduction of systems that extract the explosive event agent, upon its
detection, and prior to its detonation/ignition
5. The elimination of ignition sources
6. Management of access control, for both persons and vehicles, to preclude
entry and accidental initiation of the explosive event
7. The introduction of explosive event pressure relief features that direct the
effects of the explosive event in a manner that provides acceptable postevent effects
8. Informed placement of potential explosive systems and/or environments so
that their post-event effects meet the project’s security performance criteria
9. The use of appropriate signage and graphics to prevent occupancies, activities, and/or behavior that could trigger the event
10. The use of hardening to preclude unacceptable and/or disproportionate
damage from the event
These and other mitigating strategies should be considered for implementation
by the design team to enhance the facility protection profile.
2.6.2 Intentional Explosions
This subsection addresses purposeful extraordinary criminal activity associated
with the use of explosives as a means of attacking persons and property. Refer
to Section 2.3, “A Brief History of Recent Terrorist Attacks,” which identifies
sources of related explosive event history and demographic information. While
engineering design guideline publications can assist the design professional by
providing pertinent related historical information, ownership and design professionals must anticipate that history will repeat itself and that the design and
client community have a responsibility to engage in predictive security planning.
THREAT REDUCTION
61
This handbook advocates the use of the extensive explosive event information available from numerous agencies, such as the FBI, the Department
of State, the Department of Defense, the Department of Homeland Security, the U.S. Armed Forces, and others, as a means to reinforce confidence
in providing—well beyond the publication date of this manual—protection
against explosive events.
While the primary use of explosives was in the military arena, criminals and
terrorists have adopted the use of explosive devices as the most commonly used
tactic to inflict devastating results. Contemporary with the publication of this
handbook, law enforcement agencies, the intelligence community, terrorism experts, and security professionals continue to emphasize that the manportable improvised explosive device (MPIED) and vehicle-borne improvised explosive device (VBIED) are the weapons of choice for the criminal
to use when contemplating an attack against a population and/or facility,
with the intention of achieving mass loss of life, disproportionate property
damage, and the inevitable publicity this generates.
Facility designs outside of North America are also subject to the use of explosives as part of the criminal/terrorist act modality. In addition to the MPIED and
VBIED delivery methods, the use of standoff military ordnance, especially the
use of shoulder-fired arms (such as the rocket-propelled grenade [RPG]) remains
a significant concern and should be given careful consideration by a structural engineer and architect as part of a competent protective design mitigating strategy.
It is especially important, for work outside of North America, that design
professionals insist upon the use of a competent threat and risk assessment
process that is specific to the project facility, in order to determine whether
standoff military weaponry is to be considered as part of the mitigation
criteria.
It is also essential for the design professional team, especially the structural
engineer and architect, to understand the relationship among procurement of explosives, their delivery, placement, and subsequent detonation, in order to clearly
understand the rational and design consequences associated with the identification of design basis threats. The protection strategies and eventual design implementation concepts and components of a facility’s structure, facade, ERR systems, and business-mission continuity are founded upon mitigation of either a
predetermined explosive event type, size, and detonation location, or a predetermined pressure, impulse, and post-event facility performance standard. Consequently, the establishing of quantifiable design basis threats is a fundamental first step in providing the necessary guidance to the design team, as this
will be the basis for subsequent analysis and design mitigation development.
Explosive design basis threats (DBTs) must be provided as part of the threat
and risk assessment by a competent security professional. The project-specific
and identified DBTs will be based upon the opportunity for the aggressor to procure the explosive, the ability for it to be delivered to the point of attack, and
the opportunity to achieve a detonation sequence that will achieve the full value
of the DBT value. Threat and risk assessment professionals, working with the
62
DESIGN CONSIDERATIONS
law enforcement and intelligence communities, have resources to assist in the
development of DBT values based on these generic parameters. However, it is
essential for the design professional team to participate in the risk assessment
process by the provision of project-specific information, at the earliest stages
when conceptual designs are adequately advanced. This is necessary as it will
inform the ability for explosive delivery mechanisms to be considered credible or
not credible, based on project design features that affect the mobility of vehicles
and persons intent on delivering the VBIED and MPIED, respectively. While
the threat and risk assessment authors can provide quality information regional to the site, the extent to which the project site and facility physical
plant are at risk from specific explosive event DBTs must be informed by an
interactive process of information exchange between the threat assessment
and design team professionals.
Large quantities of explosives are readily delivered by vehicles. These values are considered public sector information-sensitive. However, there is a trend
toward this information becoming common knowledge within the design community as more explosive event protection mitigation is practiced. However, this
handbook continues to respect the need for this information to remain confidential. Consequently, the explosive delivery capabilities for passenger-operated
vehicles (cars); passenger vans, sport utility vehicles, and pickup trucks; small
box trucks; large box trucks and trucks with vessels carrying liquids; and the
larger classes of vehicles such as tractor trailers have all been assigned explosive
event design basis threats by the law enforcement community. Quality sources
of VBIED DBT information are the ATF, TSWG (Technical Support Working Group), and other DHS agencies. Design professionals should request
this information from the project’s security professionals engaged in the
project threat and risk assessment.
MPIED DBTs have been credibly established as a result of their extensive use
in numerous criminal acts. Again, there is a range of explosive event DBT sizing
based on the delivery mechanism. Individuals may deliver explosives either by
transporting them directly as a body-worn device or as a separately carried parcel
or satchel. The increased use of well-packaged high explosives allows for competent concealment and provides the law enforcement and security professional
with significant challenges for rapid identification through the routine use of informed visual observations. There are emerging programs that employ behavioral identification, parcel recognition, and other predictive strategies employed
under the nomenclature “situational awareness.” These are supplemental procedures only; they are helpful in explosive device identification, but should not be
considered substitutes for physical protective designs.
The body-worn explosive device DBT is not insignificant and is capable
of inflicting structural damage. The extent of this damage can be dramatically increased through the use of high-energy explosives and specific charge
shaping.
The satchel may be carried or packaged in a wheeled parcel pulled by the aggressor. Wheeled parcels often are harder to detect because the ease with which
VULNERABILITY REDUCTION
63
they can be transported reveals less through body language. Consequently, the
MPIED has a potentially large value range. It is significantly in excess of the
body-worn DBT value.
Explosives can also be purposefully delivered using the U.S. Postal Service,
express couriers, common freight carriers, or other similar second- and thirdparty delivery agencies. These differ from the MPIED threat, as they arrive
cloaked by the legitimacy associated with an expected delivery mechanism and
devoid of the opportunity for profiling the carrier. The range of DBT value is
entirely dependent upon the parcel/freight weight and size management threshold criteria established by the carrier. Suffice it to say that the range of explosives that can be delivered by a carrier generally exceeds that which an aggressor
could deliver personally. With the advent of GPS and other handheld intelligent
geospatial and initiating technologies, criminals are capable of remotely detonating shipped explosives at explicit latitudes, longitudes, and elevations, allowing
targeting by building, by floor, and by room, assuming satellite signal transmission strength is adequate.
In summary, design professionals should use the security consultant’s estimates of the VBIED and MPIED threats in the design of protection strategies for the structure, facade, ERR systems, and so forth. They should insist
upon active participation in those portions of the threat and risk assessment
that involve the identification of building design features and components
that can affect the type, size, and location of a bomb. Also, a competent security systems and operations professional should work with the threat and
risk assessment and design team to advise on the method to limit the delivery of explosives.
2.7 VULNERABILITY REDUCTION
This section addresses the critical topic of how vulnerability of an institutional facility, including its physical plant (buildings), occupants, and assets
can be secured—both from the standpoint of safety and of business-mission
continuity—by the application of rational and justifiable mitigating strategies.
Facility vulnerability can be defined as the capability of, or susceptibility to, being compromised, damaged, destroyed, or otherwise violated in a manner that is
detrimental to human safety and institutional viability.
It is unreasonable to expect that government agencies can entirely eliminate
the threat of a bomb. As mentioned previously and as so aptly demonstrated
during the Second World War, the kamikaze attack or the actions of the suicide
bomber generally are successful, as they defy the basic concepts of military engagement, law enforcement, and protective designs, which assume that even the
enemy or the criminal exercises the intent of self-preservation. In an environment in which criminal threats cannot be dismissed, only diminished, institutions remain vulnerable, and protective designs that address those vulnerabilities are necessary.
64
DESIGN CONSIDERATIONS
Previous portions of this handbook also identified the potential for accidental
explosive events, which also place structures and their occupants at risk. While
these may be significantly diminished by competent planning and protective
design strategies, there always remains the opportunity for exposures resulting
from protective designs that are overwhelmed by larger threats and/or combinations of events that were not anticipated in the risk assessment.
Factors of safety (FOS) are routinely employed in the designs executed by engineering professionals, especially structural engineering, which benefits from
the values that have established industry standards for seismic, wind, and gravity loads. Unfortunately, there are no specific FOSs for intentional explosions,
and for accidental explosions there are very limited published guidelines (such
as those published by electrical utility companies for transformers, by steam utility providers for high-pressure steam distribution systems, etc.) identifying the
quantified extent of additional protection that should be designed for by the structural engineer. Consequently, the identification of facility vulnerability to an
explosive event, which in large part depends on the value of the explosive
threat to be mitigated, remains an inexact science. While the military has developed practical approaches to address military attack ordnance of particular
sizes and types, this information is not explicitly helpful in identifying the vulnerability of conventional structures to explosions associated with nonmilitary
ordnance.
The process of vulnerability reduction must therefore occur within a design
environment in which there is a presumption that the threat cannot be completely
eliminated, that its quantification is an informed but still subjective value, and
that there are no preestablished industry factors of safety. Irrespective of the
difficulty in specifically quantifying explosive events, there remain viable
opportunities and design considerations for use by design professionals to
address facility vulnerability. They are addressed in the following sections
and can be summarized as standoff distance, physical security and protective design, operational security, and structural design responses.
2.7.1 Standoff Distance
Standoff distance is the amount of space denied to an aggressor between his
threat device and the area being protected; for example, ram-resistant vehicle
barriers set 100 feet from the building facade create a 100-foot standoff. Since
the destructive effects of an explosion decrease rapidly over distance, space is
the most effective and cost-efficient safety measure of all the blast-mitigation
strategies available to building designers and engineers. (See the later chapters
on blast phenomena and blast loading for detail.)
The DoD and other federal entities7 provide uniform tables of minimum
standoff distances that the design team can use to establish and enforce their
7
See the National Counterterrorism Center (NCTC) Bomb Threat Standoff Distances table at
www.nctc.gov/docs/2006 calendar bomb stand chart.pdf.
VULNERABILITY REDUCTION
65
agreed-upon security and safety levels. Determining the correct standoff margin with precision is challenging, however, since estimating the required
standoff distance involves an accurate approximation of the bomb’s composition and explosive weight. Because the terrorists hold the initiative, such
calculations are often simply educated guesses and hence become dependent
on the accuracy of the threat assessment.
Employing standoff design principles to protect the facility’s facade and primary and secondary structure, ERR systems, business-continuity-related intellectual property, physical plant, and other protected assets requires sophisticated
planning, additional budget allocations, and probably architectural innovation or
restraint. Putting space between the explosive device and the assets remains
one of the design team’s best mitigating strategies for risk and consequence
management, so realizing the positive benefits of standoff distance should be
a primary focus of the site selection process. Consequently, vehicle anti-ram
perimeters, pedestrian interdiction barriers, and other site civil and landscape
features should be given serious consideration.
2.7.2 Physical Security
Physical security equipment and operational security measures such as personnel, policies, and procedures are proven defenses that can keep explosives away
from the facility. Although security programs are a clear and observable protective measure—think of controlled underground parking and the badges now required in most office buildings—design teams often neglect to build competent
and enforceable physical security equipment and operational security program
spaces into the facility’s infrastructure.
2.7.3 Operational Security
Architects’ and engineers’ failures to anticipate security program needs seem
rooted in their clients’ tendency to institute these measures only after building
occupants begin arriving. But any facility constructed to resist explosive attacks
needs—to cite only a few examples—controlled access areas, delineated zones of
control, and places where vehicular and pedestrian entry is strictly managed. Design professionals should vigorously pursue with ownership the opportunity
for positioning the required protective equipment and security operations
spaces into their planning cycle in a manner compatible with the intended
facility operation’s plan. The final design will yield a more secure facility with
greater functional capability for detection, deterrence, prevention, and recovery.
2.7.4 Structural Design
This subsection provides a nontechnical discussion of how structural design
should be considered as one of the major mitigating strategies for reducing facility vulnerability. Since the analytical and engineering specifics of this topic
66
DESIGN CONSIDERATIONS
are addressed in much greater detail in other sections later on in this handbook,
this subsection is intended to introduce and validate the role of explosive-eventresponsive structural designs. This section also identifies the significance associated with making the determination that structural robustness is to be part of the
facility’s protective design strategy, along with the accompanying effects on the
design process, product, construction cost, and risk acceptance.
The decision to enhance a facility’s structural design to resist the effects of explosive events and thereby reduce its vulnerability will have project management
design implications. These include:
1.
2.
3.
4.
5.
Owner decisions regarding risk tolerance and acceptance
Project cost allocations
The effects on the role of structure in architectural design
Design space implications for architectural use programs
Design implications for mechanical, electrical, and plumbing (MEP) and
other building infrastructure systems
6. Implications associated with construction quality control for both structural shop drawing submittals and during construction field inspections
7. The requirements imposed on document production and management to
maintain confidentiality of the security-sensitive information included in
the construction document packages
Assuming that standoff distances are optimized through a competent security
program, which achieves maximum available separation between the explosive
DBT and the structure, and that a combination of physical and operational security strategies achieves credible reductions to agreed-to explosive DBTs, remaining vulnerabilities from the explosive threat will need to be managed by
the facility’s structure. Facility protective designs are dependent upon reducing project-specific vulnerabilities, and the extent of mitigation that must
be performed by the structure is the residual threat not capable of being
addressed by separation from the threat and the extent of threat that is
not capable of being detected and precluded from access to the site and the
facility.
Even small explosive events create extraordinary load magnitudes in excess of
wind loads on facility facades. These applied loads to the facade must be transferred back to and managed by the building’s structural frame. Additionally, explosive events, based on the project-assigned DBT values, will very likely apply
significant loads directly to both primary and secondary structural members and
elements, such as columns, girders, trusses, shear walls, connections, and slabs.
This may come about as a result of larger DBT values at some distance from the
structure (externally), or these larger DBTs may have proximal/adjacent access
to the structure itself. Structural exposure to and the consequent vulnerability to
these threats can occur from either the VBIED or MPIED, as noted previously.
VULNERABILITY REDUCTION
67
Additionally, these same VPIED and/or MPIED threats may internally threaten
the structure, depending on the project occupancies and the extent and efficacy of
the facility’s security programs in managing the migration of the threat. In summary, the structural frame/load-carrying system of the building will likely
provide the most significant protective value in reducing vulnerability to an
explosive event, as the loss of primary and secondary structure will usually create circumstances that place significant numbers of building occupants; evacuation, rescue, and recovery systems; and business missions at
severe risk.
However, with the decision to engage the structure in facility vulnerability reduction, one of the most expensive building components is now subject to modification in order to provide performance beyond that required by most building codes and conventional wind, gravity, and seismic loading. While European
building codes (such as those in the United Kingdom) currently embrace the requirement for design professionals to address the removal of a single column
(for unspecified reasons, essentially requiring the designers to address a “threatindependent” vulnerability), this requirement has not yet been uniformly adopted
by building codes throughout North America. And, when this requirement has
been included in a particular regional building code, or as a programmatic design
conformance requirement, the removal of a single column may not provide the
extent of protection required to preclude progressive collapse and disproportionate damage from threats of specific sizes as identified within the project threat
and risk assessment DBTs. These DBTs (used in “threat-specific” or “threatdependent” designs) may be of such a size and capable of achieving adjacency
to the structure so that they affect the load-carrying capability of more than a
single primary element. These circumstances, therefore, require a greater extent of mitigation and consequent vulnerability reduction than would otherwise
be provided by the use of the threat-independent criteria only. Consequently,
a fundamental risk management decision must be made by project ownership and codified in the key programmatic design and facility performance
assumptions to be used by the design professionals: specifically, whether the
structural system for the building will provide protection against the actual
project threat assessment identified explosive DBTs (“threat-dependent” design approach), or the extent of protection afforded by addressing the removal of a single primary structural element (“threat-independent” design
approach), or, potentially, both criteria. The client-defined extent of postexplosive-event facility performance must be clearly understood and agreed
to in order for there to be a clear expectation regarding structural performance in vulnerability reduction. This may have a significant project cost
impact, and, during the initial Conceptual phases of the project’s design,
magnitudes of probable cost should be assigned to this aspect of vulnerability reduction and a determination made regarding funding viability and
assignment. Inability to fund a competent protective structural design response
will inevitably lead to a commensurate increase in vulnerability and an increase
in the owner’s risk exposure.
68
DESIGN CONSIDERATIONS
Based on the project explosive event DBT values and the locations where
these threats can migrate within the project site and structure, a wide range of
structural protective design strategies may emerge. This may involve the introduction of additional structural members, increasing the size of structural elements, the introduction of new structural framing and load-carrying techniques
such as trusses and outrigger systems, addition and diversity of building materials, connection detailing, and other schemes, all of which may affect the overall appearance of the structural framing systems. Based on the aesthetic role
of structure in the particular project’s architectural vocabulary, there may be a
direct visual translation of these protective design influences on the architect’s
visual design intent. During the initial Conceptual phases of the project’s design, blast consultants and structural engineers should participate in design
charrettes with the project architects. These charrettes should address the
symbiotic relationship between the structural scheme options and architectural design intent to achieve the required blast protection and the desired
structural component of the architectural concept. Failure to address this
vulnerability reduction opportunity at this early phase within the project
may prejudice the opportunity to develop the most architecturally acceptable, security-competent, and cost-effective structural scheme.
While visual aesthetics often dominate design decisions, structural designs
responding to the requirement for blast mitigation may involve potentially significant impacts on programmatic space use, size, and arrangement. Increased
member sizes for blast protection, the addition of structural members or systems,
and/or the introduction of physically enforced standoff to improve structural survivability will, of necessity, reduce space available for other functional purposes.
Also, large structural systems such as trusses, additional shear walls, compression and tension rings, outrigger systems, and other similar systems may be of
such size and/or space invasiveness as to dictate relocation of architectural space
programs and/or MEP systems to floors other than where these structural systems need to occur to provide the required levels of protection. These space constraints may not be the only motivation to reassign architectural use programs or
MEP systems to alternate locations within the system. Based on the facility security program, the introduction of the threat and risk assessment identified DBTs
may place these critical structural systems at an enhanced level of risk. This occurs because it may not be possible to preclude (due to the space utilization and
occupant activity programs) the MPIED or VBIED from having access to the
desired locations where these structural system enhancements optimally would
occur. Consequently, during the initial Conceptual phases of the project’s
design, the blast consultant, structural engineers, and security professionals should participate in design charrettes with the project architects so that
there is agreement upon the ability to feasibly implement a security program
that precludes the explosive DBTs from having access to critical portions of
the structure. Also, these sessions should resolve the identification of structural impositions on MEP systems’ design and deployment, and the anticipated space impacts, both in terms of space availability and occupancy use
VULNERABILITY REDUCTION
69
programs. Failure to address these structural vulnerability reduction opportunities, at this early phase within the project, may prejudice the opportunity to develop the most competent and cost-effective structural schemes,
architectural designs, space utilizations, and MEP systems’ placement.
Value engineering has become an inherent project management process
throughout the design and construction profession. It has been shown to provide
beneficial results by optimizing the engineers’ design intent through a process
involving the application of previous experiences through a peer review by other
design professionals, contractors, and/or material fabricators. As long as this process continues, opportunities exist to modify structural systems in a manner that
may or may not reduce their ability to provide the enhanced protection they must
provide when they are used as part of an explosive event facility vulnerability
reduction strategy. The science of designing structural systems to resist explosive event loading is a unique and less well understood art than the management
of wind, gravity, and seismic loading. It is extremely likely that those involved
in the value engineering process may not fully understand the extent of blast
load structural system investment included by the original design engineers, and
recommendations for materials and/or structural systems’ design optimization
may be based upon inadequate knowledge or experience. Any value engineering and/or other similar “design optimization” process that involves review
and recommendations regarding structural systems engaged in blast vulnerability management should include the blast consultant and the structural
engineers responsible for designing the systems under evaluation.
Furthermore, during the construction phase of activity, the structural engineers
should remain involved in the shop drawing approval process to ensure that the
materials’ fabrication, erection, and field installation techniques and details continue to respond to the analytically proven designs that were developed as part
of the original blast mitigation work to achieve the required project vulnerability
reductions to the threat-assessment-identified explosive DBTs. Value engineering and quality assurance involvement during the construction phase by the
blast consultant is essential. As a result of the blast consultant’s analysis
and recommendations to protect against an explosive event, the structural
materials’ selection and the systems’ design and detailing may well be, of
necessity, more robust than that otherwise required for traditional gravity,
wind, and seismic loadings. Without the benefit of this blast analysis, the
rationale for unconventional material oversizing, encasement, overly robust
connections, and other structural enhancements may appear to be simply
overly conservative structural design and detailing, and may thereby create
an intuitive but false opportunity for invalidated optimization by the contracting community.
Structural designs providing explosive event protection are codified within
the drawings and specifications used for both procurement and directing contractor construction-related activities. Of necessity, details associated with member sizing, connections, reinforcing detailing, plating, and other blast protection techniques must appear in adequate detail to inform bidding, shop drawing
70
DESIGN CONSIDERATIONS
development, and field work. The structural engineering professionals should
develop, with the project architect and ownership team, the means to ensure
the nondisclosure of, and maintenance of the confidentiality of, structural
enhancements intended for the purpose of reducing vulnerability to the
project explosive DBTs. This process should be in place prior to the development of any documentation, whether Conceptual designs, blast analytics, or
more detailed documentation normally associated with the Schematic, Design Development, and Construction Document phases of design work. This
should include a series of published policies and procedures to manage electronic
and hardcopy design information.
2.8 RISK ACCEPTANCE
It is universally acknowledged within and outside of the design profession that
no facility can actually be designed, constructed, and operated in such a manner
as to be totally risk-free. There are simply too many variables and unknowns that
can create exposures and vulnerabilities that cannot be foreseen or completely
mitigated. The challenge is to reach a rational and justifiable position of balance
among available resources to be applied to facility design, construction, fit-out,
and subsequent operations, and to do so based on the best contemporary knowledge available at the time of design and on reasonable foreseeable future vision.
These resources include finances, design talent, design time and schedule, site
opportunities and constraints, materials availability, contracting skill, construction schedules, post-construction operational staffing, intellectual resources to
manage operations and create and enforce viable policies and procedures, and
last but not least, client risk tolerance and commitment.
This section discusses the implications of designing facilities to meet explosive threats effectively with less emphasis on resource implications and with
the primary goal of achieving lower levels of risk. This approach is generally
considered to be “designing to address the threat” or, more simply, “design to
threat.” The emphasis being that vulnerability reduction and threat mitigation
are the leading programmatic requirements for consequence management and
are intended to dominate the protective design decision process.
This section also discusses the implications of implementing designs that include mitigating strategies as they can be afforded by available resources, and
the consequent potential increases in risk exposure if these resources are limited. This approach is generally considered to be “designing within budget” or,
more simply, “design to budget.” This remains as one of the more significant
challenges to the design professional and client team, as there are almost always limited resources available, and this then requires an eventual trade-off
between resource allocation and protective design implementation. Just as it
is unreasonable to assume that no design and facility operation will be devoid
of threats, vulnerabilities, and risks, so is it unrealistic to assume that there are
RISK ACCEPTANCE
71
unlimited resources available to achieve complete risk elimination. The process
of adjudicating this tension between resource consumption and facility security enhancements will require the design professionals to clearly identify
optional levels of protective enhancement so that client risk management
and the risk tolerance allocation process have a basis for rational decision
making.
2.8.1 Design to Threat
This subsection focuses on the process of assigning and allocating resource dedications adequate to fully address the threats identified in the threat and risk
assessment process as referenced in Section 2.4 and the associated advantages in
terms of increased opportunities for substantial risk reduction. Consequently, the
success of the design team in achieving the extent of risk acceptance identified
by ownership will, in part, be based upon a clear understanding of the projectspecific threats identified in the threat and risk assessment. As mentioned previously, for explosive events, this will require a clear and concise definition of the
details associated with the types of events. This may include purposeful criminal, man-made devices or naturally occurring opportunities for there to be an
explosion that places the facility at risk. For the criminal act, this will include the
type of explosive to be used for analytical modeling (usually, this is expressed
in terms of TNT equivalent), the size of the design basis threats (usually, this is
expressed in terms of a specific weight of TNT equivalent), the delivery mechanisms (MPIED, VBIED, etc.), and other specific characteristics (such as charge
shaping, etc.). It is important to note that a project may well have multiple DBTs
for a variety of different threat scenarios. For other explosive event opportunities,
well-defined criteria need to be established, including the source of the explosive agent, its containment or delivery vessel or system, quantities, and other
relevant specific data necessary to achieve the appropriate level of analytical
modeling.
Once the threat data are defined and the post-explosive-event facility performance criteria are fully understood and agreed to with ownership, the design
to threat approach requires the design professionals to engage in the process of
clearly identifying facility vulnerabilities. This will, of necessity, involve the services of security professionals so that the potential delivery of threats can be
clearly defined based on the architectural design, planned electronic and operational security countermeasures, and other explosive threat-reduction program
elements. This will allow a clear understanding of remaining areas of explosive
vulnerability to both criminal and noncriminal explosive threats. Since the resource investments required to substantially reduce risk from explosive events
are normally substantial, addressing the specific areas of greatest vulnerability and then identifying remaining areas of lesser vulnerability is prudent. This
process of prioritizing vulnerability from most significant to least significant
is positively dependent upon the design professionals clearly identifying, and
72
DESIGN CONSIDERATIONS
agreeing with ownership on, where an explosive threat can occur. The development of a rational and known to be competent security program, which
identifies with confidence limited areas of threat potential, is the most economical means to assign resources to limited areas of vulnerability mitigation. This will achieve the best risk reduction with the least resource
allocation.
While reducing the quantity of vulnerabilities has substantial virtues, those remaining vulnerabilities may be significant and challenging. The design to threat
approach, quite simply, requires the design professional to develop the appropriate mitigating strategies during the design process, so that the post-event performance criteria meet the standards originally established during the Conceptual
Design phase, as codified in the project key design assumptions. In anticipation
that required resource allocations will not be trivial, the design professional team
may find it most beneficial to develop two or three alternate schemes of structural system design, so that there is an inherent and preestablished response to
the formal or informal value engineering process, which will inevitably occur as
part of any well-managed project.
For the structural engineer, the design to threat process must assume that
threat reduction opportunities have been fully exploited. Simply stated, opportunities for reducing the DBTs have been taken into consideration by law enforcement agencies and by other project mitigation strategies developed by ownership and the balance of the design professionals. As an example, this may
include the provision of security personnel, equipment, and policies and procedures, such as the use of explosive detection devices and screening operations as
a means to reach an agreed-to explosive DBT level at various points throughout
the site and facility. Assumptions involving the definitive reduction of explosive DBTs and their potential delivery locations should be specifically validated by a well-orchestrated and documented series of design charrettes
involving ownership, the architect, the security professional, and the structural engineer.
The design to threat approach may, in fact, inform ownership that structural system protection, hardening, or alternate load transfer path designs create substantial resource allocation hardships. Irrespective of the development
and presentation of various structural design options intended to address
the explosive threat, it may be necessary to carefully examine the means
to reduce the threat, reduce the number of vulnerabilities, or have ownership participate in greater risk acceptance. This may occur as a result of the
unattractive and/or unmanageable financial or programmatic implications of the
structural protective design options. It is for this reason that addressing the opportunity to employ the design to threat approach during the earliest Conceptual
and Schematic design phases is essential. The viability of designing to specifically achieve the required threat reduction through a process of structural design
mitigation may identify the need to impose and implement the other threat mitigation resources. This may include the development of protective architectural
and site civil design features, deployment of security systems, equipment, and
RISK ACCEPTANCE
73
manpower, and recurring capital expenditures for operations. It may also involve
increased restrictions associated with space-use programs and/or redesign of the
systems associated with mitigating the effects of accidental explosive events.
2.8.2 Design to Budget
This subsection discusses the counterpoint to the design to threat process. Specifically, the design to budget process can be invoked by ownership when they
determine that finite resources are available and a set quantity is established;
the extent of protection provided is then tailored to achieve the best risk reduction benefit within those strictly defined resource boundaries. The term “budget”
in this section of the handbook refers to the budget of overall resources, a definition purposefully intended to be more than just project funding. However, more
often than not, the project resources that will have the most confining effect on
protection strategies intended to mitigate the explosive event will be the project
funds available to spend on security. Occasionally, other resource issues, such
as lack of specific protective design materials and/or the skilled labor to install
them, will also create resource constraints and may direct the design professional to utilize an overabundance of those protective strategies that are the most
available. As an example, the inability to include a competent electronic security
system as part of the project, whether due to lack of skilled labor, system components, or the ability for that system to be reliably maintained and serviced, may
dictate that explosive event facility vulnerabilities require greater investments
in physical hardening to make up for the inability to implement early warning
detection and control of access.
Once there is agreement on the project definition of budget resources and the
extent of funding available for an agreed-to definition of “project security,” the
design professionals should conduct a series of design charrettes. At a minimum,
these should include the project architect, structural engineer, blast consultant,
MEP and vertical transportation consultants, site civil and landscape designers, and security professionals. Other consultants, including traffic engineers,
lighting consultants, geotechnical engineers, and others may need to participate
as well, based on the project’s location, occupancy program, and complexity.
The intent in sequestering the design teams’ resources is to create a series
of consensus-developed options (at least two but preferably three) wherein
each design discipline inherits a small portion of the responsibility for mitigating the facility’s vulnerability to explosives. This has repeatedly been
shown to be the least expensive and most effective approach to resource allocation and will provide the best opportunity to meet both protective design
and budget objectives. In short, this is the most expeditious path toward “designing to budget.”
Unfortunately, these design charrettes, while yielding the most comprehensive
and effective solutions at the least cost, often still identify funding expenditures
in excess of client resources. This is usually a product of design professional conservatism, wherein each professional advances protective designs that have the
74
DESIGN CONSIDERATIONS
absolute least opportunity to fail at that professional’s particular mission. Ownership’s misallocation of funding (underfunding), due to their extremely limited
experience and understanding of the investments required to create truly credible explosive event consequence reduction designs, is also a factor. Sometimes,
however, there are simply insufficient overall project funds to create great architecture that includes spaces that are intended to support continued critical
operations in environments of considerable threat potential.
When there is a disparity between the ownership’s available resource pool
and the resource requirement dedications to meet the originally agreed-to postexplosive-event performance, the design to budget process approach does not
need to be abandoned. However, the design approaches that remain will all involve an extent of risk acceptance by ownership. Quite simply, at this point, there
is an admission by the design and ownership teams that vulnerabilities will continue to exist in the project that may not be adequately addressed. In order to
achieve the best result, the project vulnerabilities to the explosive event DBTs
identified in the threat and risk assessment will need to be listed and prioritized
based on the anticipated consequential effects of an explosive event. This can
be most successfully accomplished if there is a comprehensive and prioritized
list of consequential effects. A thorough understanding of the prioritized vulnerabilities and prioritized consequential effects will allow the available project
resources to be applied in a manner that provides the best extent of protection
and risk management. Once a determination has been made about how limited project resources are to be allocated, a concise written protective design
narrative should be authored by the design team. This narrative should be
subsequently reviewed and approved by ownership so that there is a legacy
document that clearly explains the rationale behind resource allocation and
the extent of explosive event protection provided.
It is essential for the design professionals to understand that client bracketing
of funds for project security will need to cover investments other than facade,
structural systems, ERR systems, and sensitive space hardening. Security strategies to reduce the explosive event threat will also require expenditures for site
civil vehicle approach interdiction features, electronic security systems, both personnel and vehicular screening equipment, and the costs for security staffing and
the build-out of their programmatic support spaces. Based on the project costestimating techniques and/or models used, redundant diverse ERR systems and
architectural features, and redundant diverse business continuity systems may be
lumped into the overall security budget. In order for the design to budget process to be effective—and in order to avoid mismanagement of the extent of
funding available for actual explosive event security mitigating building features, components, and systems—the design professionals must engage in a
reconciliation process with ownership. This partnership should address and
agree upon what project features and components constitute an explosive
event protective design strategy and the budget available to achieve it. This
activity should take place at the time when the project budget is in the process of
being established.
RECENT EXAMPLES OF SECURITY DESIGN “BEST PRACTICES”
75
2.9 SOME RECENT EXAMPLES OF SECURITY DESIGN “BEST
PRACTICES”
The authors’ combined 80-plus years of security experience have acquainted
them with a broad range of effective and workable protective building-design
concepts and mitigation solutions, summarized below, that are in daily use
around the world. (Because of the sensitivities surrounding post-9/11 security,
venues are not specified.8 ) Allocating space for these measures, and incorporating utility requirements for them, are considerations that security design teams
need to address at the earliest design stages.
1. Defense in Depth—Creating a layered zone defense built upon concentric
rings of interlocking physical security, security systems and equipment,
personnel, and policies and procedures.
2. Security Zones of Control—Enforcing a distinction between different physical spaces and the level of security assurance that each space must provide.
Some locations are designated as sterile; that is, everyone in them has been
screened for explosives (as well as chemical/biological/radiological materials and other hazardous or prohibited items). Other spaces are identified
as secure; that is, only people with proper access privileges are permitted
to enter. Other areas may be open to the general public and without restrictions. Additional zones of control may also exist based on the project type,
size, and location.
3. Redundancy and Diversity—Designing for the redundancy and diversity
of critical assets, including security, ERR, and MEP systems, as well as
client-specific physical and intellectual assets.
4. Exceptionally Protected Zones—Specifically designing for the preservation of unique cultural artifacts and the protection of unique financial assets, intellectual resources, critical unique industrial or financial processes,
and other single-source assets that do not allow for protection through diverse repetition. To protect one-of-a-kind assets, hardening solutions have
to be created in a manner responsive specifically to each asset’s vulnerability.
5. Personnel Screening—Identifying and screening all individuals and their
personal items for explosives and other proscribed or hazardous materials.
8
These sites include the U.S. Capitol, the Pentagon, Central Intelligence Agency (CIA) Headquarters, Boston’s Logan International Airport, the London Ring of Steel, London’s Canary Wharf commercial office complex, and the Eurotunnel between the United Kingdom and France. This best
practices section also includes observations from the Sears Tower in Chicago and the Statue of
Liberty Visitor Screening Center. Relevant security measures at a worldwide insurance company
headquartered in lower Manhattan and a number of comparable private sector enterprises, as well
as several U.S. border inspection stations, are also incorporated.
76
DESIGN CONSIDERATIONS
6. Vehicle Screening—Identifying and screening all vehicles to prevent
vehicle-borne explosive (and other) threats from entering managed street
segments, on-site parking, subgrade parking, and cargo offloading areas.
7. Cargo and Mail Screening—Ensuring that qualified and trained personnel
screen for explosives all cargo deliveries destined for the buildings’ core
interior areas in accordance with the facility-specific DBT. For example,
security personnel would inspect all overnight and freight couriers, and
other nonpostal/mail services for explosives and other threats. In some very
high-security venues, all delivery vehicles, mail, and cargo may need to be
inspected off-site before delivery to the protected premises.
8. Secondary Screening Capability—Enforcing a clearly defined separation
between the primary inspection process for screening people and vehicles
and detecting anomalies, and the secondary inspection process, which resolves any anomalies through detailed search, investigation, and testing.
At least two different detection technologies should be used to complement each other—for example, using explosive-detection canines, explosive vapor sampling, trace detection, or detailed visual inspections to conduct secondary investigations when high-energy imaging appears to detect
explosive materials.
9. Space and Provisioning for Security Control Centers—Ensuring that the
facility command, control, communications, and intelligence (C3I) center
that monitors and controls security systems and operations is afforded the
appropriate space, utilities, security, environmental protection (from smoke
and chemical/biological agents, as appropriate), and backup systems.
10. Segregation and Protection of Building Mechanical, Security, and ERR
Equipment—Isolating critical security, ERR, and MEP systems and areas
from tampering and disabling by unauthorized individuals.
11. Crime Prevention through Environmental Design (CPTED)—Designing
CPTED—a concept that incorporates environmental design features
into the overall security strategy for deterring crime and unwanted
behavior—into an area or facility.
2.10 RELATED PHENOMENA
Designing to comprehensively mitigate the explosive event requires an understanding of the consequential effects associated with the release of significant
amounts of energy over very short durations. Design professionals must be cognizant that the explosive event alone is not the only threat consequence requiring
design attention and subsequent mitigation. Post-explosive-event effects may be
numerous based on the type of explosive event, location internal or external to
the facility, the type of building construction, internal contents, occupancy use
program, and numerous other contributing factors such as the flame, smoke, and
fuel contribution of the structure itself and its contents.
RELATED PHENOMENA
77
No discussion of “related phenomena” would be complete without the acknowledgment that there will be the potential for severe injuries and fatalities;
damage to critical facility assets, both intellectual and physical; disruption of
business missions; and the need for implementing evacuation, rescue, and recovery operations. Additional effects will include crime scene management,
post-event facility evaluations, insurance claim investigations, and reconstruction
and restoration of physical plant, institutional operational viability, and public
image.
However, design professionals have the opportunity to influence a limited
venue of related explosive event phenomena. With appropriate assistance from
blast and security consultants, engineers can design to contain and manage fires
which may follow some explosions. Also, limitations to the extent of post-event
damage and structural system collapse can be analyzed and responded to, so that
the structural performance of the facility is engaged in a limited participation
in life-safety, ERR, and business continuity exposures. Well-developed facade
designs can reduce injuries and fatalities while also protecting interior spaces,
the ERR, and critical business continuity systems. Design professionals can also
identify the intended means of evacuating building populations and engage in
the planning and providing for the subsequent arrival of emergency responders
so that effective ERR can enhance the opportunity for occupant safety. In doing
so, they can clearly identify those building features, systems, and components
that require an enhanced level of explosive event protection as it is extremely
likely that they will be severely compromised by such an event unless they receive extraordinary protection.
Of the numerous explosive-event-related phenomena, the following handbook
sections emphasize the need for the design team to pay particular attention to
and dedicate both design and building resources to the mitigation of progressive
collapse, protection of ERR, and management of attendant fires.
2.10.1 Progressive Collapse
As mentioned earlier in this chapter of the handbook, the losses of primary and
secondary components of the building’s structure are potential results of an explosive event, whether a purposeful and malicious act or an accidental or naturally occurring incident. The extent of damage, as previously noted, will depend
upon the size and type of event, its location, and the extent to which the structural
system in the building has been prepared for managing the extraordinary applied
loads and any significant impacts due to airborne debris.
Since life safety is generally the paramount criterion, the definition of disproportionate damage used to define the project’s post-explosive-event security performance criteria may be dependent upon the size and location of the
building’s occupant population. High-rise office buildings with commercial occupancies traditionally allocate 150 to 200 square feet per person as a means
to estimate population densities. With common floor plate sizes ranging from
25,000 to 45,000 square feet, there are opportunities for approximately 150 to
78
DESIGN CONSIDERATIONS
300 people per floor. The loss of multiple floors, especially throughout the entire
height of the building, which may be as many as 70 or 80 floors, defines catastrophic loss of life circumstances. Hence, progressive collapse from one floor
to the next is deemed unacceptable. However, the definition of disproportionate damage may be more conservative than the prevention of damage outside
of the area of explicit and direct explosive event effects. Design professionals
should, at the earliest possible design phase in the project, review with the
architect the anticipated occupancy profiles for the building’s interior and
exterior spaces. They should examine the extent of failure associated with
the threat-assessment-identified explosive event DBTs or other possible explosive sources within the facility, should no supplemental structural robustness be added for explosive event management. At this point, a determination should be made whether the consequent resulting structural damage
invokes too high a casualty count for the owner’s and design professionals’
risk tolerance. If that is in fact the case, then while the damage may not be
structurally disproportionate, the life-safety consequences may be unacceptable.
This may then inform a more conservative definition of allowable disproportionate damage. Since the design professionals, from a structural standpoint,
generally are responsible for designing to preclude progressive collapse and
to achieve nondisproportionate damage, clear definitions of these terms, especially disproportionate damage, should be reviewed with ownership and
codified as part of the project’s key design assumptions. The significance of
these definitions becomes even more evident for low-rise facilities that have significant occupancies concentrated in relatively small areas, such as performance
venues or transportation structures. In these environments, there may be, for estimating purposes, one person for every 4 to 6 square feet, which may translate into thousands of persons within a relatively small area. Again, a traditional
structural definition of disproportionate damage may define acceptable structural
damage to four column bays as a result of the loss of a single column. However,
the loss of these bays may translate to, using a 40-foot-square bay module, 6,400
square feet of occupancy. In this example, at a density of 6 square feet per person, there are potential casualties of over 1,000 people. This is an unacceptable
result. Therefore, each project must be individually evaluated and the progressive
collapse designs developed accordingly.
In order to competently evaluate either preexisting structures intended for
renovation and/or expansion or the design of new structures, the services of a
qualified blast consultant working in partnership with the structural engineer are
essential. The determination of the extent of anticipated damage and the opportunities to identify the extent of potential mitigation will be based upon their
interactive use of structural analysis, blast effects science and software, historically informed engineering judgment, and the validation of results conclusively
drawn from these experiences and analytical tools through actual live testing of
structural systems with actual explosive events. Failure to properly anticipate
and define the circumstances and designs that will not prevent progressive
RELATED PHENOMENA
79
collapse and that will not prevent disproportionate damage will preclude the
elimination of potentially fatal flaws in the design and place both ownership
and the design professionals at considerable risk.
2.10.2 Disruption of Evacuation, Rescue, and Recovery Systems
Careful study of post-explosive-event conditions clearly identify the value provided by building designs that provide a safe and secure means of egress for
internal occupants and equally safe and secure means of reentry for emergency
responders. Time-motion egress analytics, which model the amount of time
required for a building to engage in a complete evacuation of all its occupants, routinely show dramatic improvements when all of the intended
means of egress are available and unimpaired. Or, stated another way, the
loss of egress capacity associated with an explosive event can be expected to
dramatically increase the time required for people to exit a facility at risk and to
substantially diminish emergency responder efficacy at extracting the base population of handicapped individuals and those incapacitated by the explosive event
itself. This can be expected to translate into higher injury and casualty counts, as
every moment is crucial during facility occupant extraction.
Structural integrity of stairs and elevators (and sometimes escalators, as these
may also be used as an egress mechanism in some facilities) may not translate
to a tenable ERR environment. As exemplified by the 1993 use of explosives at
the World Trade Center by terrorists, stairwells became filled with smoke, even
though they were structurally adequate to continue to function as viable means
of egress. This smoke condition hindered exiting operations and was allegedly
responsible for smoke inhalation, and a significant number of related adverse
health consequences were claimed. Structural integrity maybe inadequate if not
supported by survivable smoke control/pressurization systems combined with
interlocked and functionally survivable smoke detection.
At the time of authorship of this handbook, there is a distinct and evolving
trend for building codes and other legislative documents that guide and/or direct
the design community to require the survivability of evacuation, rescue, and recovery systems after an explosive event. Considerable attention has been focused
on this subject as a result of the extensive studies carried out by the National Institute of Standards and Technology (NIST) of the World Trade Center attacks
of September 11, 2001. Numerous building safety recommendations were made
by the NIST committee, many of which focus on a need for additional legislation and design responsiveness to the enhancement and preservation of ERR
systems, building features, and design components. It is reasonable to expect,
and this handbook anticipates, that these NIST recommendations will become
the foundation for new language in national and international building codes. It
is reasonable to assume that such significance has been focused on ERR survivability that these codes will call for increased stairwell widths; more robust
stairwell construction; the use of elevators for ERR operations, and consequently
80
DESIGN CONSIDERATIONS
the construction of more robust and water-resistant hoistways and lifts; enhanced
fireproofing ratings, adhesiveness and impact resistance; improved fire detection
and reporting; more survivable and more competent fire suppression systems;
more effective emergency voice communication systems; and other related support infrastructure improvements. With these known to be valuable and soon
to be codified ERR building features, design professionals should acknowledge that one of the most significant explosive-event-related phenomena requiring attention is the preservation of facility ERR operations.
In Section 2.5.2, a comprehensive list of ERR systems was provided. While
it is not reasonable to expect that these systems will survive in the immediate
zone of the most powerful explosive event effects, design professionals should
carefully analyze the location where the threat-assessment-identified explosive
event DBTs can occur. They should then design ERR systems so that they are
appropriately protected to achieve survivability and continue to provide functionality to the balance of the facility. This may involve protective hardening
of the spaces where they occur or the design of redundant systems with adequate geographical diversity so that they are not disabled by the DBT events.
To accomplish this effectively is no small design endeavor. A quick review of
the ERR building components will intuitively identify that they are likely to be
spread throughout the entire facility. This may therefore require a detailed review of the entire facility and all of the attendant systems and analytical modeling of the protective design features in response to all of the various explosive
DBTs. Consequently, the design professionals should carefully review the
client’s expectations, code requirements, and good engineering practice and
develop a clearly stated standard with respect to the desired performance
of ERR systems and building features in response to the project’s explosive DBTs. To do this competently, it must be remembered that many systems
that support ERR are functionally dependent upon each other. As an example,
stair pressurization, smoke purge, and smoke control fans, necessary to maintain egress and firefighting, are integrally linked to the building management
system (BMS), fire command systems, and emergency power, and the compromise of any one of those by the explosive event will render that system inoperative. Consequently, all of these systems must be systematically analyzed and
protected. As another example, fire suppression systems, necessary to manage
potential attendant fires, are dependent upon the integrity of fire suppression
piping, pumps and control panels, emergency power, and fire alarm systems.
Again, a compromise of one of these components may place the entire system
at risk.
There are simply too many permutations of ERR systems integration that
could be placed at risk by explosive events to competently note and review each
of them individually in this handbook for the purpose of developing uniquely responsive mitigating strategies. The appropriate process to ensure that ERR systems have the greatest probability of post-event survivability is further discussed
elsewhere in Chapter 2. as part of consequence management, and is founded
upon the strategy of overlaying ERR systems on the project threat maps and
RELATED PHENOMENA
81
having their designs and/or hardening protection informed by individual explosive event modeling using the project DBTs.
2.10.3 Attendant Fires
Not all explosive events are accompanied by an immediately ensuing fire development. The rapid expansion of gases at hypersonic speeds and the instantaneous
consumption of oxygen as part of the explosive event chemistry often deny fire
development. To that end, the oil industry has used oxygen starvation as a means
of suppressing oil well fires through the use of explosives. However, certain types
of explosive events, such as those caused by high-pressure gas mains, often have
an initial burst of high energy but then continue to burn at extreme temperatures and have the opportunity to subsequently engage combustible materials or
continue to generate adequate heat from their own fuel source to subject structural elements to ductile yielding and reduction in load-carrying capacity. Mitigating strategies often involve the opportunity to eliminate the participation of
secondary fuel sources ignited by an explosive event. The design professionals
should carefully review the storage and distribution systems for flammable
agents and their exposures to the project explosive DBTs and include in
their design pump shutdowns, automatic and manual breach control valves,
and/or other strategies so that these fuel sources have limited burn times.
Additionally, the design of robust encasement around fuel storage and/or
distribution systems should be considered. Examples include the installation
of generator fuel storage reserves in what are essentially concrete bunkers providing both explosive event ignition protection and elimination of secondarysource fuel supply.
Attendant fires pose a range of problems. As mentioned previously, they may
challenge the structural integrity of the facility as a result of extreme temperatures, especially applied in an environment where fireproofing may be damaged
and/or absent as a result of the initial blast pressures and abrasion/impact of airborne debris associated with the explosive event. Although it is anticipated that
building code legislation will likely require increased fire ratings for structural
members and the specification of more physically resilient fireproofing methods, it is unlikely that these material specifications will provide adequate postexplosive-event performance reliability. The structural engineer and architect
should partner with the project blast consultant and security professional
and determine the opportunity to receive enhanced benefits from the specification of robust fireproofing strategies. This may include the use of concrete or
masonry encasement. Reasonable solutions for structural system protection from
fire will continue to include fireproofing, fire suppression systems, and alternate
load path designs. However, the most effective mitigating strategy remains
the reduction and/or elimination of the explosive threat, and the reduction
of building materials and contents that have the opportunity to participate
as fuel sources in the fire development triad of oxygen, fuel load, and an
ignition source.
82
DESIGN CONSIDERATIONS
The National Fire Protection AssociationStandards has quantified the parameters by which building occupants are affected by fire events. Temperature, air
toxicity levels, and visual obscurescence are all factors that affect either health
or the ability to extricate oneself from the fire event. Explosive events that are
accompanied by attendant fires are likely to be characterized by significant quantities of airborne debris, irrespective of the smoke contribution associated with
the fire. Air quality may rapidly approach lethal thresholds as a result of escalating air temperatures (140◦ F is the upper-level threshold for human exposure). Air
quality may also become untenable as a result of toxic air particle concentration
from the explosive demolition effects and/or from the fire’s oxygen consumption
or toxic gas generation. These conditions combined with visual obscurescence
create a potentially deadly environment. This situation and the previous concern
regarding structural integrity are exacerbated by the likely failure of sprinkler
or gaseous suppression systems in the immediate area of the event. However,
containing the event-based fire to a smaller radius of effect will do much
to improve life safety for those outside of the explosive event zone. Consequently, the design of sprinkler systems with smaller fire control zones,
increased water application density, fast response sprinkler heads, and survivable water supplies provides a very reasonable and competent means to
both improve life safety and reduce property damage. As part of any explosive event protection strategy, the blast consultant and the structural engineer
should partner with the fire suppression engineer for the purpose of optimizing
the sprinkler system’s response to an explosive event. This should include sprinkler coverage zone sizing; the use of automatic and/or manual breach control
valves, in conjunction with alternate main and/or pipe riser cross connections,
to ensure system performance outside of the explosive event zone; increased
hazard classifications to improve water availability and discharge density; head
selection; and other pertinent aspects of suppression systems’ design with the
intention of fire containment to a limited area. This will have the dual benefit of
reducing loss of life and property and reducing the extent of structure exposed to
extreme heat.
There may be less opportunity to influence the selection of the actual building
construction materials and the interior fit-out of furniture, fixtures, and equipment, to reduce the building’s internal fuel load, but, as an explosive event design
mitigation strategy, it should not be summarily overlooked. The less there is to
burn, the shorter the duration and overall exposure of the building’s occupants
and structural elements to the deleterious effects of a fire born out of an explosive
event. Similar design logic follows in the selection of building construction materials. The specification of noncombustible components will reduce fire damage.
Generally, structural systems and building contents are regulated by the Building
and Fire Codes with the intent to reduce combustibility. However, whenever the
project threat assessment identifies explosive events as a threat to be mitigated, building and interior fit-out components should be specified with an
eye toward reducing the facility fuel load.
SECURITY DESIGN CONSIDERATION GUIDELINES
83
2.11 SECURITY DESIGN CONSIDERATION GUIDELINES
The following bullets summarize major points in the building security design
considerations process in a condensed format.
1. During the project’s design’s initial Conceptual phases, the blast consultant and structural engineers should participate in design charrettes with
the project architects so that the symbiotic relationship between the structural scheme options and the architectural design concepts to achieve the
required explosive event protective designs are considered for potential implementation. Failure to address this vulnerability reduction opportunity
at this early phase may prejudice the opportunity to develop the most architecturally acceptable, security-competent, and cost-effective structural
scheme.
2. The entire security design team should participate in the creation of a brief
design mission statement, which clearly and concisely states the client’s
expectations for the facility and site’s post-explosive-event performance.
Since these expectations are routinely overstated by clients—as a result of
their lack of sophistication or knowledge of the engineering complexities,
architectural program compromises, post-construction institutional operational constraints, impact on project financials, aesthetic effects, and other
factors better understood by the structural engineer and the balance of the
design team professionals—this mission statement must be a consensusdeveloped document.
3. On projects where explosive event consequence management is a programmatic design response requirement, design professionals should insist that
the threat assessment and site-specific DBT be provided to them at the earliest possible date. This is essential if the specific site and facility vulnerabilities are to be identified and mitigating concepts developed for potential
implementation.
4. During the site selection process, the security and blast consultant explosive event engineering professionals should work closely with the
client as full members of the building security design team. Based on
the maximum realistic standoffs, the pressures and impulses that will
be applied to the structure, its facade, inbound utilities, and other critical site civil features should be factored into the site selection decision. This is essential so that an initial determination can be made on
how these extraordinary loads (in excess of gravity, wind, and seismic) will affect the protective design capabilities of the major building
systems.
5. The initial phases of consequence evaluation and management should occur in the Conceptual and Schematic Design phases and include a review
84
DESIGN CONSIDERATIONS
6.
7.
8.
9.
10.
of the opportunities for redundancy and diversity and the use of hardened
facility structural, facade, and ERR design elements.
Insufficient efforts to clearly define the client’s consequence management
expectations are one of the most significant design process shortfalls and
are a major reason why the engineering efforts fail to meet client needs or
expectations.
Deliberate or accidental explosive events rarely impact only the areas defined within contract boundary limits of work, a fact that must be taken into
consideration as part of consequence evaluations. During the initial stages
of site planning and facility design blocking, stacking, and facade selection, consequence evaluations should consider the aggressors’ opportunity
to deliver the explosive to the target and the extent of damage a detonation
external to the building or site could cause.
Design professionals must address building designs so that the explosive
event consequence management response is based on an evaluation of how
ERR will occur. In this way, designs are informed to protect ERR building
design features and provide inherent tactical support and definitive criteria
for an ERR concept of operations.
While the pressures, impulses, and plasma effects associated with an explosive event using extremely high-energy substances such as plastic explosives are regarded as upper-level protection design thresholds, the security
design team should carefully review the entire project design to identify
other building systems or storage spaces that could participate in an explosive event.
Any value engineering or other similar “design optimization” process
that involves review and recommendations regarding structural systems
engaged in blast vulnerability management should include the blast
consultant and the structural engineers responsible for the systems under
evaluation.
2.12 CONCLUSION
It is the authors’ hope that this discussion of security design considerations has
given the reader an increased awareness of how a competent and well-organized
security design approach, initiated at the beginning of the design process, can
significantly influence the eventual success or failure of the facility’s explosive
event protective design strategies. Managing the design and security interdependencies and complexities of today’s blast-resistant construction, in an uncertain
and ever changing threat environment, requires the contributions of security, explosives, and building sciences domain experts working together with ownership
risk managers, forming a group that can then appropriately be called the security
design team.
REFERENCES
85
REFERENCES
Boettcher, Mike and Ingrid Arnesen. 2002. Al-Qaeda Documents Outline Serious
Weapons Program. CNN.com (January 25, 2002), available at http://www.isisonline.org/publications/terrorism/cnnstory.html.
Federal Emergency Management Agency. 2005. Risk Assessment: A How-To Guide
to Mitigate Potential Terrorist Attacks (FEMA 452). Washington, DC: Federal
Emergency Management Agency, Department of Homeland Security. Available at
http://www.fema.gov/plan/prevent/rms/rmsp452.shtm.
Robertson, Nic. 2002. Bomb-making video reveals scope of al Qaeda threat. CNN.com
(August 21, 2002), available at http://archives.cnn.com/2002/US/08/21/terror.tape.
main/index.html.
Santayana, George. 1905. The Life of Reason, or the Phases of Human Progress: Introduction and Reason in Common Sense. New York: Charles Scribner’s Sons.
U.S. Department of Defense. DOD Dictionary of Military Terms. http://www.dtic.
mil/doctrine/jel/doddict/.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
3
Performance Criteria for
Blast-Resistant Structural
Components
Charles J. Oswald
3.1 INTRODUCTION
Performance criteria specify quantitative limits on the response of structural
components that are intended to achieve the blast design objective or strategy for
the building design, such as those summarized in Table 3.1. A structural component is designed by calculating the blast load on the component, using a dynamic
analysis to determine the maximum component response, and then checking the
response against quantitative limits that are consistent with the overall building
design objective.
In typical static design, the stress level of component response is limited to
prevent failure. However, this approach cannot be used for blast design, since
controlled, ductile yielding is usually part of the design intent. Therefore, response limitations are typically placed on the maximum dynamic deflection of
the component to prevent component failure, where this deflection accounts in
an approximate manner for the amount of acceptable damage or plastic strain.
The amount of conservatism in the design depends on the allowable component deflection limits. Usually, the allowable component deflection is limited so
that either repairable, or unrepairable, damage occurs, but not component failure.
In other cases, when component failure from a worst-case explosion scenario is
almost impossible to prevent, response limitations may be placed on the postfailure velocity of the components and the amount of overall building collapse.
Table 3.2 shows different areas of blast design and typical performance goals
for each area. The performance goals are affected primarily by the type of blast
threat for each area. For example, terrorist attacks involve many uncertainties related to the explosive threat, and consequently this area of blast design has more
basic design performance goals. Explosive safety, on the other hand, involves explosions of known stored or manufactured materials, and its performance goals
have more defined levels of protection. Hardened military structures typically
contain mission-critical equipment and personnel (i.e., command and control
87
88
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.1
Summary of Blast Design Strategies
Performance Goals
Design Strategy
Comments
Prevent component
failure
Limit response of components
protecting personnel and
equipment to levels below those
causing failure. A safety factor
against failure is provided by
minimizing expected damage
levels.
This is the primary approach used
to protect building occupants
against injuries from an
explosion.
Limit structural
collapse
Prescribe more conservative
allowable response criteria for
load-bearing components. Also,
design buildings to withstand
localized failure of a primary or
load-bearing component (i.e.,
design against progressive
collapse).
UFC 4-023-03 (UFC 2005) and
GSA (GSA 2003) have specific
criteria to design against
progressive collapse. Building
collapse will almost always
cause a very high percentage of
fatalities to building occupants.
Maintain building
envelope
Design against failure of building
cladding elements, including
windows and doors.
Blast overpressure entering
buildings can cause significant
injuries and extensive damage,
but not usually many fatalities.
Minimize flying
debris
Design cladding components using an Mitigating the extent and severity
of injuries to occupants in
expected worst-case explosive
building areas without overall
threat to fail without becoming
structure collapse is very
hazardous projectiles to the
dependent on minimizing flying
occupied space. This may be used
debris.
for a limited building area receiving
the worst-case blast loads.
Prevent cascading
explosion events
Design structures around stored or
manufactured explosives to resist
projectiles and blast pressures from
a nearby explosion without failing
catastrophically and detonating
explosives within structure. Also,
design building components to
protect personnel and control
equipment necessary for shutting
down potentially explosive
surrounding industrial processes in
an explosion.
This includes proscriptive
requirements based on testing in
DDESB 6055.9 (DDESB 2004)
and blast design requirements to
protect equipment in UFC
3-340-02 (UFC 2008) for cases
involving high explosives and
ASCE (ASCE 1997b) for cases
involving industrial explosions.
centers) that must be protected against a wide array of enemy threats, but these
threats are relatively well defined.
3.2 BUILDING AND COMPONENT PERFORMANCE CRITERIA
Blast design and analysis are primarily component-based, whereby the applied
blast load and response of each component in the building are determined
BUILDING AND COMPONENT PERFORMANCE CRITERIA
89
Table 3.2 Summary of Blast Design Areas
Design Area
Performance Goals
Comments
Explosive safety
Protect personnel
Protect equipment,
supplies, and stored
explosives Prevent
cascading explosive
events
Design against accidental explosion where
the amount and location of potential
explosives are relatively well quantified.
All inhabited areas are designed to
prevent injuries. This type of blast design
generally has the most design
conservatism.
Hardened military
structures
Preserve
mission-critical
functions
Buildings designed to resist attacks from
specified weapons with a near-miss or
direct hit such that given operations can
still be performed.
Antiterrorism
Prevent mass casualties
Minimize flying
debris Limit
structural collapse
There are uncertainties in the amount and
location of potential explosions caused
by terrorists. The usual intent is to build
in a given amount of blast resistance
using a level of protection against a given
defined threat. A maximum standoff
distance to the protected building is
established with barriers and inspections
for explosives.
Process safety
Protect personnel
Prevent cascading
events Minimize
financial losses
Explosive events can be modeled relatively
well, but initial conditions of the accident
causing explosion must be assumed.
Neutral risk philosophy is employed such
that people inside buildings should have
the same protection as those outside
buildings. Ability to shut down
operations is often important to prevent
cascading events.
individually. However, performance goals are usually set in terms of life safety,
functionality, and reusability for the entire building. Therefore, damage or response levels for individual building components must be established to achieve
the overall building performance goal. A framework for this process has been developed by the American Society of Civil Engineers (ASCE, forthcoming) based
on a similar approach developed by the U.S. Army Corps of Engineers, Protective Design Center (PDC) (Protective Design Center 2008b). In this approach, a
building Level of Protection is selected based on desired performance goals, as
shown in Table 3.3. Table 3.4 is then used to determine the associated damage
levels for each type of component in the building. Therefore, this approach determines component damage levels that are consistent with an overall building
performance goal. It can also be used to assess the Level of Protection provided
by an existing building, based on calculated component damage levels.
Table 3.4 refers to three types of components: primary, secondary, and
nonstructural. Primary components are members whose loss would affect a
90
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.3
Level of
Protection
Building Levels of Protection
Building Performance Goals
Overall Building Damage
I (Very low)
Collapse prevention: Surviving occupants
will likely be able to evacuate, but the
building is not reusable; contents may not
remain intact.
Damage is expected, up to the
onset of total collapse, but
progressive collapse is
unlikely.
II (Low)
Life safety: Surviving occupants will likely
be able to evacuate and then return only
temporarily; contents will likely remain
intact for retrieval.
Damage is expected, such that
the building is not likely to be
economically repairable, but
progressive collapse is
unlikely.
III (Medium)
Property preservation: Surviving occupants
may have to evacuate temporarily, but
will likely be able to return after cleanup
and repairs to resume operations;
contents will likely remain at least
partially functional, but may be impaired
for a time.
Damage is expected, but
building is expected to be
economically repairable, and
progressive collapse is
unlikely.
IV (High)
Continuous occupancy: All occupants will
likely be able to stay and maintain
operations without interruption; contents
will likely remain fully functional.
Only superficial damage is
expected.
number of other supported components and whose loss could potentially affect
the overall structural stability of a building area. Examples of primary structural
components include columns, girders, and any load-bearing structural components such as walls. Secondary components generally are supported by a primary
framing component and can fail without creating widespread structural damage.
Table 3.4
Expected Component Damage for Each Level of Protection
Component Damage Levels
Level of Protection
Primary Structural
Components
Secondary Structural
Components
Nonstructural
Components
I (Very low)
Heavy
Hazardous
Hazardous
II (Low)
Moderate
Heavy
Heavy
III (Medium)
Superficial
Moderate
Moderate
IV (High)
Superficial
Superficial
Superficial
Hazardous damage—The element is likely to fail and produce debris.
Heavy damage—The element is unlikely to fail, but will have significant permanent deflections such that it is not
repairable.
Moderate damage—The element is unlikely to fail, but will probably have some permanent deflection such that
it is repairable, although replacement may be preferable for economic or aesthetic reasons.
Superficial damage—The element is unlikely to exhibit any visible permanent damage.
RESPONSE PARAMETERS
91
Examples of secondary structural components include non-load-bearing infill
masonry walls, metal panels, and secondary steel framing members such as
studs, girts, purlins, and joists. Nonstructural components include interior
non-load-bearing walls, and architectural items attached to building structural
components.
Nonstructural components are not usually designed specifically against design
blast loads. In general, their performance has received very little attention, since
testing and research for blast-resistant design has focused on structural components. However, when appropriate, depending on the level of protection, these
nonstructural components can be attached to structural members with strengthened connections in manner similar to that used in earthquake-resistant design.
3.3 RESPONSE PARAMETERS
Component blast damage levels, such as those in Table 3.4, have been correlated
to the maximum dynamic deflection of the component with two nondimensional
parameters: the ductility ratio (µ) and support rotation (θ). The reasons for basing the correlations on these simplified parameters are discussed in this section,
as well as the limitations of this approach. The support rotation is defined in
Figure 3.1:
2xm
θ = tan−1
L min
where: Lmin = is the shortest distance from a point of maximum deflection to a
support
It is based on the shortest distance from a support to the point of maximum component deflection (Lmin ), where the point of maximum deflection is determined
from yield line theory. Lmin is equal to one-half the span length of one-way spanning components, except for cantilevers where Lmin is the whole span length.
The yield line pattern for two-way spanning components can be determined from
charts in UFC 3-340-02 (Unified Facilities Criteria Program 2008), which is an
update of TM 5-1300 (DOANAF 1990).
The ductility ratio is defined as the ratio of the maximum component deflection to the component yield deflection, as shown in Equation 3.1. The yield
L
Lmin
θ = Support Rotation
XM
Figure 3.1 Support Rotation Angle
92
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
rU
Ke
Kep
1
re
KE
RESISTANCE
Ke
RESISTANCE
rU
1
1
1
Xe
Xe
Xm
XE
Xp
Xm
DEFLECTION
DEFLECTION
Determinate Boundary Conditions
Determinate Boundary Conditions
Figure 3.2 Resistance-Deflection Curves for Flexural Response
deflection for determinate components is the deflection when the stress in the
maximum moment region equals the yield strength. The yield deflection for indeterminate components is usually equal to the equivalent yield deflection, XE ,
in Figure 3.2. This is an average ductility ratio, which is less than the ductility ratio at the initial yield location and greater than the ductility ratio at the
final yield location. It is calculated as shown in Equation 3.2, such that the area
under the resistance-deflection graphs for the simplified elastic-perfectly plastic
resistance-deflection curve yielding at XE (i.e., dashed curve in Figure 3.2) and
the actual, multi-stiffness resistance-deflection curve (i.e., solid curve in Figure
3.2) are equal at deflections greater than the deflection at which the component
becomes a mechanism (i.e., Xp in Figure 3.2).
µ=
Xm
Xy
where: µ = ductility ratio
Xm = maximum component deflection
Xy = deflection causing yield of component
= Xe for determinate components (see Figure 3.2)
= XE for indeterminate components (see Figure 3.2)
re
X E = Xe + X p 1 −
ru
ru
KE =
XE
(3.1)
(3.2)
Given values of ductility ratio and support rotation have traditionally been
used as response criteria in blast-resistant design, where these criteria are the
quantitative limits that define the upper bounds of each response level or damage
level for blast-loaded components. These two parameters are used as the basis for
response criteria because they are related to the amount of component damage,
as discussed below, and because of their design-level simplicity. They are both
RESPONSE PARAMETERS
93
based on the maximum dynamic component deflection, which is relatively easy
to measure in component blast tests and to calculate with simplified dynamic
analysis methods used for blast design.
Damage occurs in blast-loaded components as plastic strains approach the
material failure strain at some location in the component. The ductility ratio is
an approximate measure of component plastic strain based on the assumption
that the curvature in the maximum moment regions increases proportionally
with deflection after yielding, and plane cross sections remain plane. In this
case, the ductility ratio is a measure of the ratio of the total strain to the yield
strain in the maximum fibers. This approach is applicable to cases where the
strains causing the component damage occur at the same location where initial
yielding occurs, such as the extreme fiber of the maximum moment region of
a steel beam. In reinforced concrete, on the other hand, damage is primarily
controlled by compression strains as they approach a given limit value (i.e., the
concrete crushing strain), whereas initial yielding occurs in the tension steel
at a deflection that usually causes relatively low compression strains in the
maximum moment region. Therefore, the ductility ratio does not correlate well
with flexural damage to reinforced concrete members. Also, very ductile steel
members typically develop tension membrane response at high ductility levels,
which can causes high strains and failure around the connections, rather than at
midspan, where initial yielding occurs.
The support rotation correlates better to component blast damage than the
ductility ratio in these cases. In the case of reinforced concrete and masonry components, the support rotation is related to the cross section rotation at maximum
moment regions, and therefore to the compression face strain, which primarily
causes damage and failure to underreinforced components. This is the case when
the component responds as a ductile mechanism after yielding of the reinforcing
steel. Very little damage occurs prior to yielding of the reinforcing steel.
In the case of tension membrane response, the support rotation increases proportionally with the maximum deflection-to-span ratio and therefore increases
with the amount of tension membrane response. Support rotation limits help ensure that the amount of tension membrane force in a ductile component that
can undergo large deflections is limited to values that do not allow connection
failures. Large deflections of primary components can also cause failures at connections to supported components. Many blast tests and field investigations at
accidental explosion sites have shown that steel components connected to their
supports with screws, bolts, and welds typically fail at the connection, rather
than at the points of maximum strain in the component (i.e., yield locations).
Therefore, ductile steel components typically have separate response criteria in
terms of ductility ratio, to limit plastic strains at the maximum moment region,
and in terms of support rotation, to limit strains from tension membrane response
near connections or other critical areas (i.e., utility cutouts in well-attached metal
studs).
The ductility ratio is usually most useful for determining damage levels with
low ductility ratios in the range of 1.0 to approximately 3.0. This applies to low
94
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
damage levels for ductile components and low to high damage levels of relatively
brittle components, such as wood components. Damage from brittle, nonflexural
response modes, such as shear, buckling, or connection failure, can also be correlated to the ductility ratio if the yield deflection, Xy , in these cases equals the
maximum component deflection causing a critical shear, buckling, or connection
strain in the component and these strains increase proportionally with increased
component deflection.
Limit support rotations and ductility ratios that establish component damage
levels generally cannot be derived from theoretical considerations alone. These
parameters are simplified for practical reasons and failure modes are complex
and influenced significantly from actions, such as combined bending, shear, and
axial load, which are difficult if not impossible to quantify and characterize universally. Also, applicable damage levels typically have qualitative, rather than
quantitative, descriptions. Therefore, limit values for these parameters are based
primarily on empirical correlations between observed damage levels to components from blast tests and corresponding values of support rotations and ductility
ratios calculated from the component’s measured maximum dynamic deflection.
This is discussed more in the next section.
A limitation in this approach is that the support rotation and ductility ratio
terms do not capture all the parameters that are actually related to the observed
component blast damage. For example, an analysis of data from fifteen oneway-spanning reinforced masonry walls tested in a shock tube showed that the
relationship between the measured support rotations and the observed damage
levels was strongly dependent on the wall reinforcement factor, which compares
the reinforcing steel ratio to the balanced steel ratio, as well as to the support
rotation (Oswald and Nebuda 2006). The test walls with a low reinforcement
factor sustained significantly less damage at the same support rotation than walls
with a higher factor. Also, a statistical study of several hundred static reinforced
concrete beam and slab tests showed that the support rotation at failure was not
constant, as generally assumed for blast design, but was a function of several
parameters, including the reinforcing steel ratio, the shear span to depth ratio,
and the ratio of the steel reinforcement yield strength to concrete compression
strength (Panagiotakos and Fardis 2001). Therefore, if two concrete panels with
different spans, thicknesses, and reinforcement ratios are both subject to blast
loads causing the same given support rotation (e.g., 2 degrees), it is possible that
they will not have the same observed damage level. A similar situation is possible
for other component types. Fortunately, this is alleviated to some extent by the
qualitative nature of damage levels, which encompass relatively broad ranges of
damage.
In spite of the limitations of the support rotation and ductility ratio as
definitive measures of component damage, it is the judgment and experience
of engineers who have developed response criteria relating these parameters to
component damage that they are adequate and generally conservative for blast
design and analysis purposes. More accurate terms that include more properties associated with a component’s blast damage level, and have acceptable
simplicity for design-based methodologies, may be developed in the future.
RESPONSE PARAMETERS AND DAMAGE CORRELATIONS
95
3.4 EMPIRICAL CORRELATIONS BETWEEN RESPONSE
PARAMETERS AND COMPONENT DAMAGE
Support rotations and/or ductility ratios that define the damage levels of blastloaded components have been established using correlations between these two
parameters calculated from available test data and observed damage levels in
the test components. The amount of available test data varies with component
type, since blast testing is typically conducted by various government agencies
and organizations to address specific blast design considerations, and the data
include varying amounts of measured material properties, blast load information,
and response values. Also, there is an almost complete lack of blast test data
for some component types and damage levels. Deflection-controlled static tests
that provide information on component performance at post-yield deflections
are considered for component types with limited available blast test data. These
static data are generally conservative because each strain and deformation level
exists for a sustained period of time, and can therefore result in more damage
than the same strain level that will only exist from a blast load for a few tenths
of a second, or less.
The most comprehensive published study that has correlated blast test data
with component damage levels is part of the Component Explosive Damage Assessment Workbook (CEDAW) methodology (Protective Design Center 2008a).
CEDAW is a fast-running methodology to assess component blast damage based
on pressure-impulse (P-I) diagrams that was developed to be as consistent as
possible with available blast test data. In this study, data were collected from
over 300 blast tests on 10 different component types and used to calculate the
support rotation and ductility ratio for each test, using single-degree-of-freedom
(SDOF) dynamic analyses. Then, this information was used to determine the rotations and ductility ratios that bounded each observed damage level for each
component type, using scaled P-I diagrams.
This effort is illustrated in Figure 3.3 and Figure 3.4, which are based on
damage levels defined in Table 3.5. Pbar and Ibar in the graphs are the peak
positive phase pressure and impulse, respectively, scaled (i.e., divided) by
relevant component dynamic response parameters to create dimensionless terms
that are directly related to component response. Therefore, components with
very different blast loads and structural properties, but the same Pbar and Ibar
values, should have the same ductility ratio or support rotation values (i.e.,
a strong component with a high blast load and a weaker component with a
lower blast load that both have response with the same ductility ratio or support
rotation would have the same Pbar and Ibar values). This should be true in both
blast tests and dynamic response calculations. The Ibar terms for component
response in terms of ductility ratio are different from those for component
response in terms of support rotation.
Figure 3.3 shows a scaled P-I diagram with curves of scaled blast loads
calculated with SDOF analyses causing ductility ratios (µ) of 1, 3, 6, and 12 for
a corrugated steel panel. These curves are plotted with data points of scaled blast
loads causing damage levels ranging from Superficial to Hazardous Failure to
96
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
100
Superficial
Moderate
Heavy
Haz. Failure
µ=1
µ=3
10
Sup. Data
Mod. Data
Heavy Data
H.Fail Data
µ=6
Pbar
µ = 12
1
0.1
0.1
1
10
100
Ibar 1
Figure 3.3 Scaled P-I Curves in Terms of Ductility Ratio vs. Scaled Data for Flexural
Response of Corrugated Steel Panels
corrugated steel panels in blast tests. Figure 3.4 is a similar scaled P-I diagram
with curves of scaled blast loads calculated with SDOF analyses causing support
rotations (θ) of 3, 6, and 10 degrees plotted against the same data that are
scaled by Pbar and Ibar terms based on component support rotation. In both
of these figures, the curves with the lowest support rotation or ductility ratio
correspond to the lowest damage level (i.e., Superficial), and curves with higher
support rotations and ductility ratios correspond to higher damage levels up to
Hazardous Failure.
Figure 3.3 and Figure 3.4 were used to determine that the ductility ratios and support rotations shown on the figures bounded each of the damage
Pbar
100
Moderate
Heavy
Haz. Failure
Mod. Data
Heavy Data
H.Fail Data
Blowout Data
θ=3
θ=6
10
θ = 10
1
0.01
0.1
0.1
1
Ibar 2
Figure 3.4 Scaled P-I Curves in Terms of Support Rotation (Degrees) vs. Scaled Data
for Flexural Response of Corrugated Steel Panels
RESPONSE PARAMETERS AND DAMAGE CORRELATIONS
97
Table 3.5 Component Damage-Level Descriptions
Damage Level
Component Damage
Blowout
Component is overwhelmed by the blast load, causing debris with
significant velocities.
Hazardous
Failure
Component has failed, and debris velocities range from insignificant to very
significant.
Heavy
Component has not failed, but it has significant permanent deflections,
causing it to be unrepairable.
Moderate
Component has some permanent deflection. It is generally repairable, if
necessary, although replacement may be more economical and aesthetic.
Superficial
No visible permanent damage.
levels, and therefore could be used as response criteria for corrugated steel panels (Protective Design Center 2008a). Similar curves were generated for all other
component types where there were available test data. The CEDAW Workbook
(Protective Design Center 2008a) has figures similar to Figure 3.3 and Figure
3.4 for each component type and a full description of the test data. Table 3.6
shows a summary of the response criteria determined for each component type
that was developed in this manner. No criteria are shown for Superficial Damage,
since this damage level was correlated to a ductility ratio of 1 by definition for
the CEDAW study. No test data were available to develop response criteria for
component types not shown in Table 3.6, including hot-rolled steel beams and
primary framing components.
PDC TR-08-07 (Protective Design Center 2008a) and Oswald and Nebuda
(2006) have a detailed discussion of the Pbar and Ibar terms used to develop
the empirical correlations between component performance levels and support
rotations and ductility ratios. These terms were developed for component
response in terms of support rotation and ductility ratio and for different types
of response modes (i.e., flexure, tension membrane, etc.) based on conservation
of energy equations between the work energy and kinetic energy from the blast
load and the strain energy of the component in the applicable response mode.
The parameters included in the Pbar and Ibar terms vary based on the type of response mode for the component and the type of response parameter representing
the component response (i.e., ductility ratio or support rotation). Also, the Pbar
and Ibar terms were developed to account for both positive and negative phase
blast loads since both types of blast loads generally affect component response
and damage levels. Extensive checks were performed on the validity of the Pbar
and Ibar scaling terms, as summarized in PDC TR-08-07 (Protective Design
Center 2008a). The scaling terms are not exact since they were derived based
on some simplifying approximations, but numerous comparisons showed they
caused the SDOF-based scaled P-I diagrams to be within 20% of more exact
solutions in almost all cases.
98
8
4
2
1.5
One-way reinforced masonry
One-way or two-way
unreinforced masonry1
4
5
2
Reinforced concrete beam
Response criteria are only valid for SDOF analyses with brittle flexural response and axial load arching.
1
Steel column (connection
failure)
1
6
Reinforced concrete column
(shear failure)
Wood stud wall
5
2
6
Heavy
One-way or two-way
reinforced concrete slab
3
Moderate
6
3
12
Haz Failure
3
2
6
Heavy
From data
w/o
SDOF
15
10
10
10
10
Haz Failure
Support Rotation
Open-web steel joist
Cold-formed girt and purlins
3
Moderate
Ductility Ratio
Response Criteria for Upper Bound of Each Damage Level for Each Component Type
One-way corrugated metal
panel
Component
Table 3.6
3
4
Moderate
6
12
Heavy
10
20
Haz Failure
Support Rotation w/ Tension
Membrane
RESPONSE CRITERIA DEVELOPMENT
99
3.5 RESPONSE CRITERIA DEVELOPMENT
The information in Table 3.6 represents the most extensive and formalized use
of blast test data to develop response criteria. This information was an important
part of the development of several of the current blast response criteria, However,
it was not the only consideration. Each organization that has developed response
criteria defining given performance levels considers a range of factors, such as
the exact definitions of its performance levels relative to the available data, how
to address component types where the data are inadequate, whether to discount
some data or data sets as not applicable, and the level of conservatism to be
included in the response criteria. There is very limited published information on
these considerations or on the compilations of the data that were considered.
Table 3.7 shows a summary of response criteria that have been developed for
each area of blast design. Ideally, there would be one general set of response limits on component response applicable for all the performance levels and design
strategies in Table 3.1. There is strong movement in this direction, but currently
there are separate sets of response limits associated with performance levels established by the different organizations associated with each area of blast design.
These differences are due in part to the fact that these organizations have different missions, design objectives, and types of explosive threats.
The following subsections provide some detail on the criteria in Table 3.7. In
all cases the response criteria are stated in terms of either ductility ratio, support
rotation, or both terms, as discussed in Section 3.2. In cases where response criteria for both terms are provided, the overall component damage level or performance level is controlled by the response criterion that causes more severe damage or lower performance. In addition to the response criteria discussed below,
the engineer should also always consider any design-case-specific requirements
on component deflection and limitations of response criteria, as discussed in Section 3.6. It should also be noted that all the response limits stated in this document
may be updated in the future. Historically, response criteria have been updated
as more information on component response to blast loads has become available
and as new component and material types are used for blast-resistant design.
3.5.1 Explosive Safety Criteria
Blast-resistant design criteria are included in UFC 3-340-02 (Unified Facilities
Criteria Program 2008), which has been developed by the explosive safety design
community under direction of the U.S. Department of Defense Explosive Safety
Board (DDESB). Explosive safety refers to safe design and operation of manufacturing and storage areas for high explosives that can accidently explode. This
primarily includes military and military contractor facilities with explosives, but
it also includes industries such as fireworks and commercial explosives manufacturers.
Most blast-resistant design prior to 1990 was for explosive safety and critical military structures, and the majority of the initial blast testing of structural
100
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.7
Summary of Response Criteria Publications
Organization
Publication
Type of Criteria
Comments
Department of
Defense
UFC 3-340-02
(Unified
Facilities
Criteria
Program 2008)
Design criteria for
explosive safety.
Published in draft form in 1984
and finalized in 1990 as the
tri-service manual TM
5-1300, NAVFAC P397, and
AFM 88 22. This document
generally has the most
conservative criteria. It has
been updated recently with
some revisions as UFC
3-340-02.
Task Committee
on Blast
Resistant
Design of the
Petrochemical
Committee of
the Energy
Division of
ASCE
Initial performance criteria for
Performance criteria
Design of Blast
blast design from 1997.
for three component
Resistant
These criteria will be updated
damage levels.
Buildings in
in a new edition of this
These criteria are
Petrochemical
publication in 2009.
commonly used for
Facilities
blast design for
(American
industry related to
Society of Civil
process safety.
Engineers
1997b)
U.S. Army Corps
of Engineers,
Protective
Design Center
(PDC)
Single-Degree-ofFreedom
Structural
Response
Limits for
Antiterrorism
Design, PDC
TR-06-08
(Protective
Design Center
2008b).
ASCE Blast
Protection of
Buildings
Standards
Committee
Based very strongly on criteria
Blast Protection of Performance criteria
in PDC TR-06-08 from U.S.
for four component
Buildings
Army Corps of Engineers.
damage levels.
(American
These criteria can be
Society of Civil
used for a wide
Engineers,
range of blast design
forthcoming)
except explosive
safety.
Performance criteria
for four component
damage levels
associated with
building levels of
protection. These
criteria are
commonly used for
design against
terrorist threats.
Performance criteria developed
in 2006 that are significantly
more comprehensive than
previous criteria. Also, they
are based more explicitly on
available blast test data.
components against conventional (i.e. non-nuclear) explosives was conducted in
support of explosive safety design. Since that time, an increase in terrorism and
industrial safety requirements by OSHA have considerably increased the amount
of blast-resistant design.
Response criteria for blast resistant design related to explosive safety for the
U.S. Department of Defense are defined in UFC 3-340-02 for the component
types shown in Table 3.8. Inhabited buildings subject to blast pressures from
RESPONSE CRITERIA DEVELOPMENT
Table 3.8
101
Response Criteria for Explosive Safety Design from UFC 3-340-02
Protection Categorya
1
2
µMAX
MAX
µMAX
MAX
Reinforced concreteb
Conventional slabs
Beams with stirrups (closed ties)
Exterior columns
—
—
3
1◦c
2◦
—
—
—
—
8◦
8◦
—
Precast concrete
Prestressed
Non-prestressed
Compression members
1
3
1.3
2
2
—
—
—
—
—
—
Reinforced masonryb
One-wayd
Two-wayd
—
—
1◦
2◦
—
—
10
—
10
2◦
2◦e
2◦
20
—
20
12◦
—
12◦
Open-web steel joists
Controlled by maximum end reaction
Otherwise
1
4
1◦
2◦
—
—
—
—
Cold-formed steel floor and wall panels
Without tension membrane action
With tension membrane action
1.75
6
1.25◦
4◦
—
—
—
—
Element Type
Structural steel
Beams, purlins, spandrels, girts
Frame structures
Plates
—
Note: Where a dash (—) is shown, the corresponding parameter is not applicable as a response limit.
a
See Table 3.9 for definitions of Protection Categories.
b
Reinforced with conventional steel rebar.
c
May be increased to 2◦ with either shear reinforcement or tensile membrane action.
d
Non-reusable walls.
e
Relative sidesway deflection between stories is limited to H/25.
accidental explosions must resist the blast loads with response criteria for Protection Category 1 in Table 3.9. Buildings protecting critical equipment must
resist the blast loads with response criteria for Protection Category 2. The component damage associated with each Protection Category is not well defined in
UFC 3-340-02, but Protection Category 1 is considered to provide a factor of
safety against failure of at least two.
The performance criteria in Table 3.8 were first published in 1984 in a draft
version of TM 5-1300 and then finalized in 1990. They are unchanged in the
recent conversion of TM 5-1300 into UFC 3-340-02. These criteria are more
conservative than more recent criteria discussed in the next sections, which are
based in part on many recent blast tests on structural components that were not
available at the time the response criteria for TM 5-1300 were developed. It
102
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.9 Protection Category Descriptions for Structural Components from
UFC 3-340-02
Protection Category
Description for Structural Components
Category 1
Attenuate blast pressures and structural motion to a level consistent
with personnel tolerances
Category 2
Protect equipment, supplies, and stored explosives from fragment
impact, blast pressures, and structural response
is also the intent of TM 5-1300/ UFC 3-340-02 to provide design criteria with
a significant level of conservatism, since it is used to help protect workers at
explosive production and storage facilities, who spend a significant amount of
time exposed to the risk of accidental explosions.
3.5.2 Response Criteria for Antiterrorism
The Protective Design Center (PDC) of the U.S. Army Corps of Engineers published Single-Degree-of-Freedom Structural Response Limits for Antiterrorism
Design in 2006 (i.e. DoD Antiterrorism response criteria) and updated it in
2008 (Protective Design Center 2008b). These response criteria have also been
adopted into the forthcoming ASCE document “Blast Protection of Buildings”
(American Society of Civil Engineers, forthcoming). The PDC response criteria
define limit values for ductility ratio and support rotation corresponding to four
different damage levels that apply to twelve different common structural component types with a total of 38 different subcategories based on component characteristics, such as response mode and types of connections. They also consider
the cases where components are primary components (i.e., major framing components) or secondary components (i.e., cladding components). These response
criteria are used to design DoD facilities against explosive terrorist threats and
can also be used for non-DoD government buildings, such as those that provide
blast protection required by the Interagency Security Committee (ISC) Security
Criteria (Interagency Security Committee 2004).
The DoD Antiterrorism response criteria are shown in Table 3.10. Some of
the criteria are based on the response limits developed from blast test data acquired to develop CEDAW (shown in Table 3.6) for corresponding component
types. The response criteria for other component types are based primarily on
static test data, as summarized in Table 3.11. The response criteria based on static
test data are generally more conservative than those based on blast test data, as
discussed previously. For example, there are limited blast test data that indicate
that conventional metal stud wall systems can resist significantly higher loads
than those calculated with SDOF analysis using the response limits in Table 3.10
(Grumbach et al. 2007).
103
—
1
1
1
1
0.7
4
1
1
0.7
1
0.5
Masonry
Unreinforcedc,e
Reinforceda
Structural steel (hot-rolled)
Beam with compact sectionf
Beam with noncompact sectionc,f
Plate bent about weak axis
Open-web steel joist
Downward loadingg
Upward loadingh
Shear responsei
Cold-formed steel
Girt or purlin
Stud with sliding connection at top
—
—
—
—
—
—
—
1◦
—
—
—
—
—
—
1
0.7
0.8
0.8
—
—
MAX
1
1
µMAX
Prestressed concrete
Slab or beam with ωp > 0.30
Slab or beam with 0.15 ≤ ωp ≤ 0.30
Slab or beam with ωp < 0.15 and without shear
reinforcementb,c
Slab or beam with ωp < 0.15 and shear reinforcementb
d
Reinforced concretea
Single-reinforced slab or beam
Double-reinforced slab or beam without shear
reinforcementb,c
Double-reinforced slab or beam with shear reinforcementb
Element Type
Superficial
Table 3.10 DoD Response Criteria for Antiterrorism Design
—
0.8
—
1.5
0.8
3
0.85
8
—
—
—
0.8
0.25/ωp
0.25/ωp
—
—
—
µMAX
Heavy
3◦
—
3◦
—
—
2◦
3◦
—
—
0.9
2
0.9
12
1
20
—
—
—
1◦
1.5◦
2◦
0.9
0.29/ωp
0.29/ωp
—
4◦
—
1◦
1◦
—
—
µMAX
2◦
2◦
MAX
Moderate
10◦
—
6◦
—
—
10◦
—
6◦
4◦
8◦
2◦
—
1.5◦
1.5◦
6◦
5◦
5◦
MAX
Expected Component Damage
1
3
1
—
—
25
1.2
40
—
—
—
1
0.33/ωp
0.33/ωp
—
—
—
µMAX
(Continued)
20◦
—
10◦
—
—
20◦
—
12◦
8◦
15◦
3◦
—
2◦
2◦
10◦
10◦
10◦
MAX
Hazardous
104
j
3
3
Blast doors
Built-up (composite plate & stiffeners)
Plate (solid)
1
1
—
—
—
—
—
—
10
20
2
1
1
3
—
1.8
µMAX
6
6
—
—
0.5◦
3◦
1◦
1.3◦
MAX
Moderate
20
40
3
2
2
6
—
3
µMAX
12
12
—
—
2◦
6◦
4◦
2◦
MAX
Heavy
Expected Component Damage
—
—
4
3
5
10
—
6
µMAX
—
—
—
—
5◦
12◦
8◦
4◦
MAX
Hazardous
Note: Where a dash (—) is shown, the corresponding parameter is not applicable as a response limit.
a
Reinforced with conventional steel rebar.
b
Stirrups or ties that satisfy sections 11.5.5 and 11.5.6 of ACI 318 and enclose both layers of flexural reinforcement throughout the span length.
c
The ASCE” Blast Protection of Buildings” document states that these response limits are applicable for flexural evaluation of existing elements that satisfy the design requirements
of Chapters 6 through 8 but do not satisfy the detailing requirements in Chapter 9 of the document, and shall not be used for design of new elements. This note does not appear in
the DoD Response Criteria for Antiterrorism Design.
d
Span to thickness ratio greater than 5. Also, reinforcement index ωp = (Aps /bd)(fps /f c ). Also see Equation 3.3.
e
Values assume wall resistance controlled by brittle flexural response or axial load arching with no plastic deformation; for load-bearing walls, use superficial or moderate damage
limits to preclude collapse.
f
Limiting width-to-thickness ratios for compact and noncompact sections are defined in ANSI/AISC 360.
g
Values assume tension yielding of bottom chord with adequate bracing of top chord to prevent lateral buckling.
h
Values assume adequate anchorage to prevent pull-out failure and adequate bracing of bottom chord to prevent lateral buckling.
i
Applicable when element capacity is controlled by web members, web connections, or support connections; ductility ratio for shear is equal to peak shear force divided by shear
capacity.
j
Also applicable when studs are continuous across a support.
k
Requires structural plate-and-angle bolted connections at top, bottom, and any intermediate supports.
l
Panel has adequate connections to yield cross section fully.
m
Typically applicable for simple-fixed span conditions.
n
Limited to connector capacity; includes all standing seam metal roof systems.
o
Values shown are based on very limited testing data; use specific test data if available.
1
0.5
0.5
1
1
1
MAX
Superficial
µMAX
Wood◦
Stud connected at top and bottom
Stud with tension membranek
Corrugated panel (1-way)with full tension membranel
Corrugated panel (1-way) with some tension membranem
Corrugated panel (1-way) with limited tension membranen
Element Type
Table 3.10 (Continued)
RESPONSE CRITERIA DEVELOPMENT
Table 3.11
105
Bases for DoD Antiterrorism Response Criteria
Component Type
Primary Basis for Response Criteria
Reinforced concrete slabs
Response limits from CEDAW study based on several blast
test series on simply supported wall panels where damage
occurred primarily in maximum moment region.
Reinforced concrete beams
Blast test data for slabs assumed to apply to reinforced
concrete beams.
Hot-rolled steel beams
Static test data.
Cold-formed steel panels and
girts/purlins
Response limits from CEDAW study based on several blast
test series on wall systems with cold-formed beams and
corrugated steel panels where lower damage levels
occurred primarily in maximum moment region,
characterized by local buckling of compression flange, and
higher damage levels including failure occurred at the
connections due primarily to in-plane reaction forces from
tension membrane response that occurs at large deflections.
Cold-formed metal studs
Static test data.
Open-web steel joists
Downward response limits from CEDAW study based on one
test series.
Wood
Response limits from CEDAW study on two test programs on
wood stud walls. One static and one dynamic test series
both indicated that typical plywood panels did not create a
fully composite section that added significantly to ultimate
resistance of stud wall system.
Prestresssed concrete
components
Static test data.
Reinforced masonry
Several test programs with shock tube and high-explosive
tests.
Unreinforced masonry
Several test programs with shock tube and high-explosive
tests documented in PDC TR-08-07 (2008a) and Wesevich
et al. (2002).
3.5.3 Response Criteria for Blast-Resistant Design of
Petrochemical Facilities
The Task Committee on Blast Resistant Design of the Petrochemical Committee
of the Energy Division of ASCE published the manual Design of Blast Resistant
Buildings in Petrochemical Facilities in 1997 (American Society of Civil Engineers 1997b), and an update will be finalized in 2009. This manual has three levels of component response to blast loads, as shown in Table 3.12, and provides
response criteria in terms of maximum ductility ratio and/or support rotation
criteria for each response level. Separate criteria are defined for different component types (i.e., steel beams, concrete slabs, etc.), and the criteria address primary and secondary components, frame components with and without significant
axial load, and shear controlled response for reinforced concrete components.
106
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.12 Component Damage Level Descriptions
Damage Level
Description
High
Component has not failed, but it has significant permanent deflections,
causing it to be unrepairable.
Medium
Component has some permanent deflection. It is generally repairable, if
necessary, although replacement may be more economical and
aesthetic.
Low
Component has none to slight visible permanent damage.
In general, the response limit values are quite low for components in which shear
or compression is significant compared to flexural response. Where adequate
shear capacity is provided, much larger response limit values are associated with
each response or damage level.
The updated version contains new response criteria that are based on a combination of criteria developed for the original publication and the response criteria
in PDC TR-06-08. The intent for the new response criteria was to expand the
existing criteria to consider a wider range of component types and subcategories
relevant for typical industrial buildings and to replace or modify existing response criteria with values from PDC TR-06-08 where there was a strong case to
be made based on the CEDAW data (Oswald 2008). Also, differences in damage
level definitions used by the PDC and by the ASCE committee (see Table 3.12)
were considered.
Table 3.13 shows the updated response limits for steel components, and Table
3.14 shows the updated response criteria for reinforced concrete and masonry
components. The criteria values for Medium and High response of cold-formed
steel girts and purlins, reinforced concrete and masonry, and prestressed concrete
components in Table 3.13 and Table 3.14 are based on similar values for Moderate and Heavy component damage levels, respectively, in PDC TR-06-08. The
criteria for Superficial damage in PDC TR-06-08 were not used in the ASCE
response criteria because this damage level was considered more conservative
than the Low damage level in Table 3.12. Therefore, values for Low response
of these component types are approximately one-half the values for Medium response. The reinforcement index that is used as part of the response criteria for
prestressed concrete components in Table 3.14 is defined in Equation 3.3.
The response criteria for all other component types not listed in the previous
paragraph are the same, or very similar, to values developed for the original
publication of the manual. In some cases this is because the values in PDC TR06-08 are similar to those in the original publication of the manual or there were
not enough supportive blast test data to justify changes.
wp =
A ps f ps
bd p f c
(3.3)
107
RESPONSE CRITERIA DEVELOPMENT
Table 3.13 Response Criteria for Steel Components
Low Response
µMAX
Component1
θ MAX
Medium
Response
µMAX
θ MAX
µMAX
θ MAX
20
12
Hot-rolled steel compact secondary
members (beams, girts, purlins)2
3
2
10
Steel primary frame members (with
significant compression)2,3,4
1.5
1
2
1.5
3
2
Steel primary frame members
(without significant
compression)2,3,4
1.5
1
3
2
6
4
Steel plates5
5
3
10
6
20
12
Open-web steel joists
1
1
2
3
4
6
Cold-formed light gage steel panels
(with secured ends)7
1.75
1.25
3
2
6
4
Cold-formed light gage6 steel panels
(with unsecured ends)8
1.0
—
1.8
1.3
3
2
Cold-formed light gage6 steel beams,
girts, purlins, and noncompact
secondary hot-rolled members
2
3
3
12
10
6
1.5
6
High Response
1
Response limits are for components responding primarily in flexure unless otherwise noted. Flexure controls
when shear resistance is at least 120% of flexural capacity.
2
Primary members are components whose loss would affect a number of other components supported by that
member and whose loss could potentially affect the overall structural stability of the building in the area of loss.
Secondary members are those supported by primary framing components.
3
Significant compression occurs when the axial compressive load, P, is more than 20% of the dynamic axial
capacity of the member. P should be based on the ultimate resistance of the supported members exposed to the
blast pressure. See PDC TR-06-08 for detailed examples of calculation for P.
4
Sidesway limits for moment-resisting structural steel frames: low = (height)/50, medium = (height)/35, high =
(height)/25.
5
Steel plate criteria can also be applied to corrugated (crimped) plates if local buckling and other response modes
are accounted for in the analysis.
6
Light gage refers to material which is less than 0.125 inches (3 mm) thick.
7
Panels must be attached on both ends with screws or spot welds.
8
Panels are not attached on both ends (for example standing seam roof panels).
where: Aps = area of prestressed reinforcement in tension zone
b = the member width
dp = the depth from compression face to center of prestressing steel
fps = calculated stress in prestressing steel at design load
f c = the concrete compressive strength
3.5.4 Blast Resistant Doors
Table 3.15 shows recommended response criteria for design of structural components in blast doors based on a desired door performance category from Design
108
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.14 Response Criteria for Reinforced Concrete (R/C) and Masonry (R/M)
Components
Medium
Response
Low Response
Component1
µMAX
θ MAX
µMAX
High Response
θ MAX
µMAX
θ MAX
R/C beams, slabs, and wall panels (no
shear reinforcement)
1
2
5
R/C beams, slabs, and wall panels
(compression face steel reinforcement
and shear reinforcement in maximum
moment areas)
2
4
6
Reinforced masonry
1
2
5
2
1
R/C walls, slabs, and columns (in flexure
and axial compression load)3
22
2
R/C and R/M shear walls & diaphragms
3
3
3
R/C and R/M components (shear control,
without shear reinforcement)
1.3
1.3
1.3
R/C and R/M components (shear control,
with shear reinforcement)
1.6
1.6
1.6
Prestressed concrete (wp ≤ 0.15)4
1
Prestressed concrete (0.15 < wp < 0.3)4
1
1
0.25/wp
1
2
0.29/wp
1.5
1
Response limits are for components reinforced with conventional rebar or prestressing strands responding primarily in flexure unless otherwise noted.
2
A support rotation of 4 degrees is allowed for R/C components that have compression face steel reinforcement
and shear reinforcement in maximum moment areas.
3
Applicable when the axial compressive load, P, is more than 20% of the dynamic axial capacity of the member.
P should be based on the ultimate resistance of the supported members exposed to the blast pressure. Refer to
PDC TR-06-08 for detailed examples of calculation for P.
4
The reinforcement index, wp , is defined in Equation 3.3.
of Blast Resistant Buildings in Petrochemical Facilities (American Society of
Civil Engineers 1997b). Table 3.10 shows similar design information from PDC
TR-06-08. Blast doors can be designed using SDOF analyses of door components responding in flexure such that the responses of all components comply
with the applicable response criteria in these tables. Both inbound and rebound
response of blast doors must typically be considered during blast door design.
The structural performance of metal doors and frames and their restraining hardware (such as latches and hinges) at the load equal to the maximum calculated
resistance from the SDOF analyses can be verified by applying an equivalent
static pressure to blast doors in accordance with ASTM F2247 (ASTM 2003b).
Blast-resistant doors are typically designed by vendors according to specifications from the design engineer and architect. Blast door performance can also
be demonstrated by blast tests. There is an ASTM committee that has developed
a draft specification for blast testing of doors (i.e., ASTM WK1902).
109
RESPONSE CRITERIA DEVELOPMENT
Table 3.15 Recommended Response Criteria for Blast Door Components
Performance
Category
Response Criteria
Description
µMAX
θ MAX
I
The door is to be operable after the loading
event, and preestablished design criteria for
stress, deflection, and the limitation of
permanent deformation have not been
exceeded. This category should be specified
when the door may be required to withstand
repeated blasts or when entrapment of
personnel is of concern and the door is a
primary exit to the building.
1.0
1.2
II
The door is to be operable after the loading
event, but significant permanent deformation
to the door is permitted. The door must
remain operable, and this category should be
specified when entrapment of personnel is a
concern
2–3
2.0
III
Noncatastrophic failure is permitted. The door
assembly remains in the opening. No major
structural failure occurs in the door panel
structure, the restraining hardware system, the
frame, or the frame anchorage that would
prevent the door assembly from providing a
barrier to blast wave propagation. However,
the door will be rendered inoperable. This
category should only be specified when
entrapment of personnel is not a possibility.
5–10
8.0
IV
Outward rebound force and resulting hardware
failure is acceptable.
Not specified
Not specified
3.5.5 Blast-Resistant Windows
Table 3.16 shows performance criteria in the form of hazard levels to building
occupants for blast-loaded windows developed in the United Kingdom and used
by the U.S. DoD (SAFE/SSG 1997). Table 3.17 shows similar criteria developed by the Interagency Security Committee (ISC 2004). These criteria are illustrated in Figure 3.5 and Figure 3.6. Response of windows to blast load is currently calculated with government-sponsored software that is not public domain,
but is available from the U.S. government for design of U.S. government buildings on a need-to-know basis. These software programs, which include Wingard
(Window Glazing Analysis Response and Design) (Applied Research Associates 2008) and HazL (Window Fragment Hazard Level Analysis) (U.S. Army
Corps of Engineers 2004), model the overall window response using an equivalent SDOF system and glass flyout after window failure for a wide variety of
110
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.16
DoD Hazard Levels for Blast-Loaded Windows
Hazard Level
Description
No break
No visible damage to the glazing or frame.
Minimal hazard
Glazing fragments inside the test structure are within a maximum
distance of one meter from the window line. For laminated windows or
windows with a film that is attached to the window frame, the window
may be cracked and fully retained by the frame, or it may leave the
frame and land outside the structure or inside the structure within 1
meter of the window opening.
Low hazard
Glazing fragments are thrown into the room for a distance of 1 to 3
meters, but do not exceed a height of 0.5 meters above the floor at the
3-meter distance. Injuries would be limited to lower body cuts, and
fatalities would not be expected, although there would be some risk to
persons within 1 to 2 meters of windows.
High hazard
Glazing fragments are thrown at high velocity into the occupied space
and impact the vertical surface at 3 meters behind the window above a
0.5 meter height. Serious injuries, including cuts to the upper body and
face from the flying fragments, would be expected. Fatalities could
occur.
Table 3.17 ISC Performance Conditions and Hazard Levels for
Blast-Loaded Windows
Performance
Condition
Protection
Level
Hazard
Level
1
Safe
None
Glazing does not break. No visible damage to
glazing or frame.
2
Very high
None
Glazing cracks but is retained by the frame.
Dusting or very small fragments near sill or on
floor acceptable.
3a
High
Very low
Glazing cracks. Fragments enter space and land
on floor no further than 1 meter from the
window.
3b
High
Low
Glazing cracks. Fragments enter space and land
on floor no further than 3 meters from the
window.
4
Medium
Medium
Glazing cracks. Fragments enter space and land
on floor and impact a vertical witness panel at a
distance of no more than 3 meters from the
window at a height no greater than 0.6 meter
above the floor.
5
Low
High
Glazing cracks and window system fails
catastrophically. Fragments enter space
impacting a vertical witness panel at a distance
of no more than 3 meters from the window at a
height greater than 0.6 meter above the floor.
Description of Window Glazing Response
RESPONSE CRITERIA DEVELOPMENT
111
Blast
HHHigh
Hazard
Zone
Window
High
Hazard
Threshold
No Break
No Hazard
Minimal Hazard
0.5 m
1.0 m
VLH – Very
Low Hazard
Zone
2.0 m
LH – Low
Hazard
Zone
Low Hazard
Threshold
Figure 3.5 Illustration of DoD Window Hazard Levels for Blast-Loaded Windows
(from ASTM 1642)
common window configurations. If the aforementioned software is not available, blast-resistant windows can be designed using an equivalent static design
approach in accordance with ASTM F2248 (ASTM 2003a) and E1300 (ASTM
2007) for a medium or low protection level. This procedure is also discussed by
Norville and Conrath (Norville and Conrath 2006).
1, 2
5
4
3a
3b
≤ 2 ft
≤ 3.3 ft
≤ 10 ft
Test window should be in the design
position or centered on the wall.
Figure 3.6 Illustration of ISC Window Hazard Levels for Blast-Loaded Windows (from
ISC, 2004)
112
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Blast-resistant windows are typically designed by vendors according to specifications from the design engineer and architect. The specifications include the
design blast load history and information on the required performance criteria.
3.5.6 Response Criteria for Equivalent Static Loads
Previously discussed response criteria based on support rotations and ductility
ratios apply for the typical blast design case where the component is designed
using a procedure that explicitly considers the dynamic component response,
such as an SDOF-based methodology or dynamic finite element analysis. Some
stiff components that are not directly loaded by blast, such as connections and
primary framing members in pure axial loading, have a short response time
compared to the expected rise time of the dynamic reaction load transferred by
supported members, and are usually designed using an equivalent static load
approach. The load capacity of the stiff component or connection must exceed
the equivalent static reaction load in these cases. If the dynamic analysis shows
the supported member does not yield, the equivalent static reaction load may be
based on the maximum elastic resistance of this member rather than the ultimate
resistance. However, this approach should be used with caution, since it will be
nonconservative if the actual blast load exceeds the design blast load.
The load capacity of the stiff component or connection is usually calculated
using applicable strength reduction factors from LRFD (load and resistance factor design) (i.e., ∅ factors), although sometimes a ∅ factor of 1.0 is used at the
discretion of the design engineer. Also, the load capacity can include a dynamic
increase factor on the yield strength, but this is typically only 1.05 for highstrength components (i.e., A325 and A490 bolts) and should be applied with
caution for cases where the capacity is controlled by a non-ductile response mode
(i.e., axial buckling or anchor bolt pullout from concrete). The equivalent static
reaction load is not multiplied by a load factor.
The equivalent static load approach has conservatism based on the use of a ∅
factor less than 1.0 (if applicable), the use of a design load equal to the theoretical
ultimate load that can be delivered by the supported component, and the fact that
a static approach is used for blast design where the maximum strains are only
applied for a few tenths of a second followed by load reversal. However, it is not
conservative if the reaction load is applied fast enough by the supported component compared to the response time of the “stiff” supporting component (i.e.
primary framing member) such that dynamic deflections occur greater than the
component’s failure deflection. This is not a concern for connections since they
are very stiff and respond quickly compared to the application of the reaction
load. Also, any component, including primary framing components, can resist
the applied dynamic load regardless of the load rise time if the component has
a load capacity greater than 1.2 times the peak dynamic load, and it can achieve
at least a ductility ratio of 3.0 without failing, based on SDOF response charts
in UFC 3-340-02 (UFC, 2008), ASCE (American Society of Civil Engineers
1997b), and Biggs (1964).
RESPONSE CRITERIA DEVELOPMENT
113
Table 3.18 Comparison of Response Criteria for Medium Damage
DoD Criteria for
Antiterrorism
ASCE Blast
Design for
Petrochemical
Facilitiesa
UFC 3-340-02b
µMAX
θ MAX
µMAX
θ MAX
µMAX
θ MAX
Reinforced concrete beam
—
2
—
2
—
2c
Reinforced concrete slab
—
2
—
2
—
1
Reinforced masonry walls
—
2
—
2
Prestressed concrete
components
—
1d
—
1d
1
2
Element Type
1
Hot-rolled steel beam
3
3
10
6
10
2
Cold-formed steel beams
—
3
3
3
10
2
Corrugated steel panelse
—
1
3
2
6
4
Open-web steel joists
—
3
2
3
4
2
Wood studs and beams
2
—
—
—
—
—
Blast door components f
10–20
6
5–10
8
a
Response criteria from draft of new edition (to be published in 2009).
Design criteria for Protection Category 1. UFC 3-340-02 does not differentiate between cold-formed and hotrolled beams.
c
With shear reinforcement.
d
Reinforcement index ωp = (Aps /bd)(fps /f c ) < 0.15.
e
Panels attached to framing at both ends with some tension membrane resistance.
f
Door not operable but remains in opening.
b
3.5.7 Comparisons of Published Response Criteria
The response criteria discussed in this chapter are compared in this section. Cases
where different published response criteria are in agreement are indicative of
more accurate correlations between the given response levels and corresponding damage or response levels, in the sense that several different organizations
agree with those correlations. Cases where there is not agreement indicate that
more blast testing is needed for these component types, damage levels, and response modes to more definitively establish applicable response criteria. It must
also be considered that it is not always possible to have direct comparisons because of differences between the definitions of response levels by the different
organizations and the stipulations associated with various response criteria.
Table 3.18 shows comparisons of response criteria for Medium/Moderate
damage from the three sources discussed in this chapter. The criteria from UFC
3-340-02 are for design of Protection Category 1 (i.e., personnel protection),
which can be considered approximately equivalent to a medium damage level.
The comparison shows component types for which there is basic agreement of
the different criteria and component types for which there is not.
114
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.19 Comparison of Response Criteria for High or Heavy Damage
DoD Criteria for
Antiterrorism
ASCE Blast
Design for
Petrochemical
Facilitiesa
UFC 3-340-02b
µMAX
θ MAX
µMAX
θ MAX
µMAX
θ MAX
Reinforced concrete beam
—
5
—
5
—
8c
Reinforced concrete slab
—
5
—
5
—
8
Reinforced masonry walls
—
8
—
5
Prestressed concrete
components
—
2d
—
2d
Hot-rolled steel beam
12
10
20
Cold-formed steel beams
—
10
12
Corrugated steel panels
—
4
Open-web steel joists
—
6
Wood studs and beams
3
Element Type
—
—
—
12
20
12
10
20
12
6
4
—
—
4
6
—
—
—
—
—
—
a
Response criteria from draft of new edition (to be published in 2009).
Design criteria for Protection Category 2. UFC 3-340-02 does not differentiate between cold-formed and hotrolled beams.
c
With shear reinforcement.
d
Reinforcement index ωp = (Aps /bd)(fps /f c ) < 0.15.
b
UFC 3-340-02 is generally more conservative than the other criteria, except
for corrugated steel panels. It is difficult to get a good comparison for this
case because the criteria in UFC 3-340-02 are for panels with “tension membrane capacity” without specifying any degree of tension membrane resistance,
as is implied in the other two criteria. The DoD and ASCE criteria agree with
each other fairly closely, except for hot-rolled steel beams and corrugated steel
panels.
Table 3.19 shows comparison of response criteria for High, or Heavy, damage from the three sources discussed in Section 3.5.1, 3.5.2, and 3.5.3.. The criteria from UFC 3-340-02 are for design of Protection Category 2 (i.e., equipment protection). All of the criteria in Table 3.19 are considered to represent
components that are at, or very near, complete failure where they would fall to
the ground or be thrown inside the building. In this case, the response criteria
in UFC 3-340-02 are not conservative compared to the criteria from the other
sources.
3.6 RESPONSE CRITERIA LIMITATIONS
A limitation of response criteria is that they are only applicable for very specific
response modes, since the amount of available ductility is a function of the mode
RESPONSE CRITERIA LIMITATIONS
115
that controls the component response for the specific case that is considered.
Unless otherwise stated, response criteria should be assumed to apply only for
components with ductile, flexural response where the component shear and connection capacities exceed maximum shear and end reaction forces, respectively,
based on the maximum resistance of the component during dynamic flexural response (i.e., equivalent static reaction forces).
This means that they are not applicable for describing damage from breach
(local failure through the full thickness in the component area of the highest blast
loads), spall (failure of the concrete near the surface in the component area of the
highest blast loads), or any other similar type of localized component response
mode. Breach and spall response are most commonly a concern for reinforced
concrete and masonry components subject to blast loads where the scaled standoff (i.e., charge standoff distance to the component divided by the charge weight
to the 1/3 power) is less than approximately 1.0 ft/lb1/3 . Several references, such
as Marchand et al. (Marchand et al. 1994), McVay (McVay 1988), and UFC
3-340-02 (UFC, 2008) have methods for predicting spall and breach in reinforced concrete slabs.
A component should be designed against blast loads so that ductile flexural
response controls the response, if possible, but if this is not possible, separate
response criteria that consider the specific type of response, such as response
controlled by shear, brittle flexure, combined flexure and axial load, or connection failure, as applicable, should be used. Unfortunately, such criteria are not
always available in published response criteria tables. Similarly, there are currently no published response criteria for many new materials that are used for
blast resistant design, such as fiber-reinforced plastic (FRP), ductile polymers,
and geotextile catch systems. In these situations, the response criteria must be
developed on a case-by-case basis by engineers with blast design and testing experience; otherwise, very inappropriate response criteria may result. For example, available response criteria for masonry walls reinforced with conventional
steel rebar have been proposed for walls reinforced with FRP, since these are
both cases of reinforced masonry, but this is very nonconservative because FRP
is much less ductile than reinforcing steel.
Another consideration is that the criteria in this chapter are intended primarily
for response calculated with SDOF analyses. However, they may also be used for
response calculated with more detailed dynamic analyses, such as multi-degreeof-freedom analysis and dynamic finite element analysis. Alternatively, design
based on dynamic finite element analysis can be based on more detailed response
criteria, such as maximum strain criteria, if the response criteria are validated
against adequate test data or are judged to be conservative enough by engineers
with significant applicable testing and blast design experience.
Finally, response criteria should also include any project-specific requirements, such as limits on deflection of a structural component from impacting
building equipment or other protected items. Another case in this category is design of blast-resistant shield walls, which are placed a few feet outside an existing
building wall to protect it from any blast loading. As the shield wall deflects very
116
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
quickly from blast load, there is a volume decrease and corresponding pressure
buildup in the annular space between the shield wall and building. The resulting
pressure can cause the conventional building wall to fail if the deflection of the
shield wall is not adequately controlled. This is not considered in the available
response criteria.
The response criteria also do not address failure of any nonstructural items attached to blast-loaded structural components, where acceptable response of wall
and roof components per the applicable response criteria may cause large enough
relative accelerations between the component and attached item to rupture the
connection. This is most critical for heavy overhead items. This subject has not
been studied very much for blast-resistant design, but earthquake-resistant design
can be used for guidance.
REFERENCES
American Society of Civil Engineers. 1997a. Structural Design for Physical Security.
New York: American Society of Civil Engineers.
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997b. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
American Society of Civil Engineers Structural Engineering Institute. Forthcoming. Blast
Protection of Buildings. New York: American Society of Civil Engineers.
Applied Research Associates. 2008. Window Glazing Analysis Response and Design
(Wingard) Computer Program, Version 5.5.1. Developed by Applied Research Associates for the U.S. Government Services Administration (GSA).
ASTM. 2003a. Standard Practice for Specifying an Equivalent 3-Second Duration Design Loading for Blast Resistant Glazing Fabricated with Laminated Glass (F2248).
West Conshohocken, PA: ASTM International.
. 2003b. Standard Test Method for Metal Doors Used in Blast Resistant Applications (Equivalent Static Load Method) (F2247) West Conshohocken, PA: ASTM
International.
. 2007. Standard Practice for Determining Load Resistance of Glass in Buildings (ASTM E1300 - 07e1). West Conshohocken, PA: ASTM International.
Biggs, J. D. 1964. Introduction to Structural Dynamics. New York: McGraw Hill Co.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5-1300, Department of the Navy Publication NAVFAC P-397, Department of the
Air Force Manual AFM 88-22). Washington, DC: Departments of the Army, the Navy
and the Air Force.
Department of Defense Explosives Safety Board (DDESB). 2004. Ammunition and Explosives Safety Standards (DOD 6055.9). Washington, DC: U.S. Department of Defense.
Grumbach S. D., C. Naito, and R. J. Dinan. 2007. Use of precast concrete walls for blast
protection of steel stud wall construction. Presented at the 78th Shock and Vibration
Symposium, Philadelphia, PA, November 4–8, 2007.
Interagency Security Committee. 2004. ISC Security Design Criteria for New Federal
REFERENCES
117
Office Buildings and Major Modernization Projects. Washington, DC: Department of
Homeland Security, The Interagency Security Committee.
Marchand K. A., S. Woodson, and T. Knight. 1994. Revisiting concrete spall and breach
prediction curves: Strain rate (scale effect) and impulse (pulse length and charge shape)
considerations. Presented at the 26th DDESB Explosives Safety Seminar, Miami, FL,
August 1994.
McVay, Mark. 1988. Spall Damage of Concrete Structures. Vicksburg, MS: Army Engineer Waterways Experimentation Station Structures Lab. Accession No. ADA199225
from the Defense Technical Institute Center (DTIC).
Norville H. S. and E. J. Conrath. 2006. Blast-resistant glazing design. Journal of Architectural Engineering 12 (3): 129–136.
Oswald C. J. 2008. Update to response criteria for blast resistant buildings in petrochemical facilities. Presented at ASCE Structures Congress 2008, Vancouver, Canada, April
24–26, 2008.
Oswald C. J. and D. T. Nebuda. 2006. Development of Component Explosive Damage
Assessment Workbook (EDAW). Presented at the 32nd DDESB Explosive Safety Seminar, Philadelphia, PA, August 2006.
Oswald C. J., D. T. Nebuda, D. Holgado, and M. Diaz. 2006. Shock tube testing on
reinforced masonry walls. Presented at the 32nd DDESB Explosive Safety Seminar,
Philadelphia, PA, August 2006.
Panagiotakos T. B. and M. N. Fardis. 2001. Deformations of reinforced concrete members
at yielding and ultimate. ACI Structural Journal 98 (2): 135–148.
Protective Design Center. 2008a. Component Explosive Damage Assessment Workbook
(CEDAW) (PDC TR-08-07). Omaha, NE: U.S. Army Corps of Engineers, Protective
Design Center.
. 2008b. Single Degree of Freedom Structural Response Limits for Antiterrorism Design (PDC TR-06-08). Omaha, NE: U.S. Army Corps of Engineers, Protective
Design Center.
SAFE/SSG (Security Facilities Executive, Special Services Group). 1997. Glazing Hazard Guide, Cubicle Standoffs, Tables and Charts, Report No. SSG/EP/4/97. London:
Special Services Group/Explosive Protection Group.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (FOUO) (UFC 3-340-01). Washington, DC: U.S.
Department of Defense, Unified Facilities Criteria Program.
. 2005. Design of Buildings to Resist Progressive Collapse (UFC 4-023-03).
Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Army Corps of Engineers. 2004. Window Fragment Hazard Level Analysis (HazL),
Version 1.2. Vicksburg, MS: USACE Engineer Research and Development Center.
U.S. General Services Administration. 2003. Progressive Collapse Analysis and Design
Guidelines for New Federal Office Buildings and Major Modernization Projects (June
2003). Washington, DC: General Services Administration.
Wesevich J. W., C. J. Oswald, M. T. Edel, M. J. Lowak, and S. S. Alaoui. 2002. Compile and Enhance Blast Related CMU Wall Test Database (CMUDS). Prepared for the
Office of Special Technology by Wilfred Baker Engineering, Inc., Contract No. 4175600-C-0900, September 6, 2002.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
4
Materials Performance
Andrew Whittaker and John Abruzzo
4.1 INTRODUCTION
The ASCE Standard Blast Protection of Buildings (American Society of Civil
Engineers 2009) permits the use of reinforced concrete, masonry, structural steel,
and timber in blast-resistant construction. Of these materials, only the two most
common materials, namely, structural steel (Section 4.2) and reinforced concrete
(Section 4.3) are discussed in this chapter. For each of these materials, constitutive models that account for strain-rate and thermal effects, procedures to account
for strain-rate effects at the macro level, mechanical properties for design, and
failure modes at the component level are discussed. Only those grades of steel,
concrete and rebar materials, and components used in the construction of typical
buildings are discussed. Steel plates, light-gage and cold-formed framing are not
structural components in typical buildings, and the interested reader is directed
to UFC 3-340-02 (U.S. Department of Defense 2008) for relevant information.
Section 4.4 discusses the use of strength-reduction factors for the computation of strength of reinforced concrete and steel components and provides the
rationale for the use of φ = 1in the ASCE Standard.
A list of key references and resource documents is provided at the end of this
chapter.
4.2 STRUCTURAL STEEL
4.2.1 Stress-Strain Relationships
Generic uniaxial stress-strain relationships are available for structural steels in
textbooks (e.g., Bruneau et al. 1997, Brockenbrough and Merritt 2006) and on
the World Wide Web. Sample stress-strain relationships are presented in Figure
4.1 (from Brockenbrough and Merritt 2006). Such relationships are established
by ASTM-standard tests and can be described using yield stress (σ y ), tensile
(or ultimate) strength (σu ), yield strain (ε y ), strain at the onset of material hardening, strain at tensile strength, and strain at rupture. ASTM specifications for
structural steels (e.g., A36, A500, A572, A913, and A992) specify minimum
values for yield and tensile strengths and strain at rupture. Limits can be set on
119
120
MATERIALS PERFORMANCE
Figure 4.1 Stress-Strain Relationships for Selected Structural Steels at Room Temperature and Low Strain Rate (Brockenbrough and Merritt 2006, with permission)
maximum yield strength and maximum ratios of yield-to-tensile-strength can be
specified for selected ASTM-designated steel. Tests to determine stress-strain
relationships, σ = f (ε), are generally conducted at low speeds and produce results at which strain-rate effects are considered negligible. Data from such relationships, after modification for strain-rate effects, are routinely used to compute
mechanical properties for design.
Rapid or high strain-rate loading affects the mechanical properties of structural steels. High strain-rate data are reported in the literature (e.g., Campbell
and Ferguson 1970, Brockenbrough and Merritt 2006, Chang and Lee 1987,
Klepaczko 1994, Lee and Liu 2006). The typical effects of increased strain rate
on the response of structural steels are an increase in yield stress; an increase
in ultimate strength, albeit smaller than for yield stress; and a reduction in the
elongation at rupture. The elastic modulus is not substantially affected by strain
rate. Sample data for selected steels at three temperatures, from Brockenbrough
and Merritt (2006), are presented in Figure 4.2.
4.2.2 Constitutive Models for Structural Steel
Finite element analysis of steel components and structures for blast effects generally involves the use of either solid or shell elements to construct a beam (column) cross section and span or a beam-column element. If solid (shell) elements
STRUCTURAL STEEL
121
Figure 4.2 Effects of Strain Rate on Yield Stress and Tensile Strength at −50◦ F, Room
Temperature, and +600◦ F (Brockenbrough and Merritt 2006, with permission)
are used to discretize a cross section, as shown in Figure 4.3, an appropriate
constitutive model must be selected for the structural steel. If beam-column elements are used to describe a structural component, the effects of strain rate and
temperature must be accommodated indirectly, as described in Section 4.2.3.
Figure 4.3 Finite Element Model of a W-Shape Steel Cross Section
122
MATERIALS PERFORMANCE
400
293°K
LOWER YIELD STRESS τ (MNm–2)
195°K
300
225°K
IV
493°K
200
713°K
373°K
II
100
I
1
102
10
∝.
0
10–4 10–3 10–2 10–1
STRAIN RATE
(SEC
103
104
105
106
–1)
Figure 4.4 Effect of Strain Rate on Mild Steel (Campbell and Ferguson 1970, with
permission)
Most of the finite element codes used for commercial blast analysis of structural components and systems include a family of constitutive models for metals
that relate stress (σ ) and strain (ε), strain rate (˙ε ), and temperature (T), namely,
σ = f (ε, ε˙ , T )
(4.1)
Some of these models are empirical, with coefficients established by curvefitting to experimental data. Empirical models seek to reproduce test data collected at alternate strain rates and temperatures, such as the dataset of Figure 4.4
for mild steel reported by Campbell and Ferguson (1970). Others are physically
based. Meyer (1992) and Meyers (1994) identify many of the constitutive models
in use today.
One of the most widely used empirical models is that of Johnson and Cook
(1983) who proposed that the basic relationship between stress and strain at low
strain rate
σ = σ0 + Aεa
(4.2)
be modified by independent terms related to strain rate and temperature, namely,
σ = (σ0 + Aεa )(1 + BIn
ε˙
)(1 − [T ∗ ]b )
ε˙ 0
(4.3)
STRUCTURAL STEEL
123
where σ0 is the yield stress, a is a hardening coefficient, A and B are experimentally determined factors, ε˙ 0 is a reference strain rate, T ∗ is a normalized
temperature calculated per Equation 4.4, and b is a coefficient determined by
curve-fitting to data.
T∗ =
T − Tr
Tm − Tr
(4.4)
In Equation 4.4, T is the temperature at which the stress is calculated, Tr is a
reference temperature at which the yield stress is measured, and Tm is the melting
point. The five parameters (σ0 , A, B, a, b) in the model of Equation 4.3 are determined experimentally. Johnson and Cook (1983), Nicholas and Rajendran (1990)
and Meyers (1994) report values of these parameters for a number of materials,
but not structural grades of steel used in buildings, bridges, and infrastructure.
The Johnson-Cook model treats the effects of strain rate and temperature independently, which may be invalid for many metals at very high strain rates (Zukas
2004, Nicholas and Rajendran 1990).
Other constitutive models are physically based and provide a relationship between stress and strain that accounts for the effects of lattice structure, grain size,
strain rate, and temperature. Two models that are implemented in LS-DYNA
(Livermore Software Technology Corporation 2003) are Zerilli-Armstrong (Zerilli and Armstrong 1987, Lee and Liu 2006) and Mechanical Threshold Stress,
both of which are described by Meyers (1994) in some detail.
The use of either empirical or physically-based constitutive models in finite
element analysis enables the user to track the effects of strain rate and temperature over the depth and length of a cross section at each time step in the analysis.
4.2.3 Component Level Strain Rate and Temperature Effects
Most applications of the Standard will involve the use of simplified nonlinear
models of structural components and indirect treatment of strain-rate effects.
Thermal effects are generally ignored because (a) most computer codes used
to compute reflected pressure histories on components do not return temperature
histories, and (b) the time frame for the conduction of heat through the depth of
a steel component is often much longer than the duration of the imposed overpressure.
The most widely used nonlinear component model has a single-degree-offreedom (SDOF). The properties of the SDOF system are computed using procedures developed by Biggs (1964) and his co-workers at the Massachusetts Institute of Technology in the 1950s. These procedures can accommodate pinned
and fixed boundary conditions and typically assume a uniform distribution of
applied pressure, which is suitable for far-field detonations only. The response
of the nonlinear SDOF system is a function of the mass per unit length, flexural
stiffness, span, boundary conditions, history of reflected pressure, and component resistance function, which is a function of strain rate.
124
MATERIALS PERFORMANCE
Consider the cross section of Figure 4.3 and assume a linear distribution of
strain across the depth of the section. For W-shape cross sections subjected to
major-axis bending, axial strain and strain rate will be greatest at the exterior
edges of the flanges and zero at the neutral axis—strain rate will vary over the
depth of the cross section. For W-shape sections subjected to axial compression
or tension, axial strain and strain rate will be uniform across the depth of the
section.
UFC 3-340-02 (U.S. Department of Defense 2008) provides guidance on
the inclusion of strain-rate effects in the computation of component axial and
flexural strength in the form of stress-increase factors. Information is provided
for ASTM A36 (carbon structural steel), A588 (high-strength, low-alloy structural steel up to 345 MPa minimum yield point), and A514 (high-yield-strength
quenched and tempered alloy steel plate) steels, for flexure (bending) and axial
force (tension or compression) and low-pressure (far-range detonation, incident
overpressure less than 100 psi, lower strain rate) and high-pressure (near-range
detonation, incident overpressure much greater than 100 psi, higher strain rate)
loadings.
The UFC presents a simple equation for strain rate, which is assumed to be
constant from zero strain to the yield strain. The equation is reproduced in Equation 4.5, below. The assumption of constant strain rate to yield is not unreasonable given the coarseness of the analysis that generally accompanies the use of
the stress-increase factors.
ε˙ =
f ds
Ete
(4.5)
where ε˙ is the average strain rate in the elastic range of response, E is the elastic
modulus for steel (29,000 ksi, 200 GPa), te is the time to yield in the cross section, and f ds is the dynamic design stress, which is a function of minimum specified yield stress, a stress-increase factor, and a factor to relate expected strength
to minimum specified strength that varies as a function of steel grade. The dynamic design stress is discussed further in Section 4.2.4
Stress-increase factors, which are termed Dynamic Increase Factors (DIF) in
the UFC and denoted by the symbol c, are used to increase the yield stress and
tensile strength of structural steels. Table 4.1, which is adapted from UFC 3340-02, presents values for c. The values of c that are applied to yield stress are
larger for flexure than axial force because blast-induced axial forces are likely
applied indirectly through beam and girder reactions for which the rise time will
be longer and the load effects staggered in time. If a component is subjected to
combined axial force and bending moment (e.g., a column subjected to bending
due to direct air-blast loading on the column flange and axial force due to indirect
loading through air-blast-induced beam/girder reactions), one value of c should
be used for component strength calculations and deformation calculations, unless
the component yield strength can be updated step-by-step in the analysis. The
chosen value of c should reflect the dominant contributor to the combined stress
125
STRUCTURAL STEEL
Table 4.1 Dynamic Increase Factors for Structural Steel (Adapted from DoD
2008)
DIF, c, for Yield Stress
Specified
Minimum
Yield Stress1
(ksi)
Low
Pressure2
High
Pressure2
Low
Pressure3
High
Pressure3
DIF, c, for
Tensile
Strength
A36
36
1.29
1.36
1.19
1.24
1.10
A5884
50
1.19
1.24
1.12
1.15
1.05
A514
100
1.09
1.12
1.05
1.05
1.00
ASTM
Grade
Flexure
Axial
1
Typical value for standard W-shape sections.
Assumed strain rate of 0.10 s−1 and 0.30 s−1 for low and high pressure, respectively.
3
Assumed strain rate of 0.02 s−1 and 0.05 s−1 for low and high pressure, respectively.
4
The values presented in the shaded cells are estimates only.
2
ratio in the beam-column. The values of c for ultimate or tensile strength are
assumed to be independent of strain rate and are close to unity.
The dynamic increase factors presented in the last column of Table 4.1 are
close to unity and could be set equal to unity with no loss of robustness in the
resultant component design. Components subjected to blast loadings that sustain
strains associated with the tensile strength of the material likely do so near the
stage of maximum displacement, where the velocity and thus strain rate are close
to or equal to zero.
It is worthwhile to compare the strain-rate multipliers in UFC 3-340-02 of
Table 4.1 with data and equations provided by others. Moncarz and Krawinkler
(1981) provide an equation for the yield strength of ASTM A36 steel as a function of strain rate in the range of 0.0002 to 0.1 s−1 , namely:
σ y = σ0 (0.973 + 0.45˙ε0.33 )
(4.6)
where σ0 is the yield stress at a (low) strain rate of 0.0002, and all other terms
have been defined previously. Using this equation, the ratio of yield stress at ε˙ =
0.0002 to ε˙ = 0.1 s−1 is 1.2. For the data of Campbell and Ferguson of Figure
4.4, the ratio of yield stress at ε˙ = 0.0002 to ε˙ = 0.1 s−1 is also approximately
1.2 (at a temperature of 273◦ K or 20◦ C).
4.2.4 Mechanical Properties for Design
Moment-curvature relationships for steel cross sections have been developed using the data from uniaxial tension tests of steel coupons, but such relationships
generally ignore the presence of residual stresses in the cross section that result from the fabrication of the rolled section (Bruneau et al. 1997). Simplified
bilinear moment-curvature relationships are generally sufficient for blast analysis computations, as described below.
126
MATERIALS PERFORMANCE
The flexural, axial, and shear strengths of steel components are typically defined in terms of cross section properties (e.g., area, web area, elastic moduli,
plastic moduli) and either yield stress or tensile strength. For blast-resistant
design in accordance with the Standard, these stresses can be increased to a
dynamic design stress by factors to account for strain-rate effects (see Section 4.2.2) and material strengths in excess of minimum specified stresses or
strengths.
UFC 3-340-02 presents a widely accepted procedure for computing so-called
dynamic design stresses. The focus of the UFC presentation is flexure. Alternate computations are made for ductility ratios of less than or equal to 10 and
ductility ratios greater than 10. Therein, ductility is defined as the ratio of maximum deflection to yield deflection. Although large lateral deflections may be
acceptable to a designer, careful consideration must be given to geometry of
the cross section at these deflections, because the cross section may no longer
be prismatic due to either element deformation due to direct air-blast loading
(see Figure 4.5) or cross section instability caused by flange (web) local buckling or lateral-torsional buckling. If the component cross section is distorted, the
design equations presented herein, in UFC 3-340-02 and the AISC Steel Construction Manual (American Institute of Steel Construction 2006), cannot be
used for component checking.
For ductility ratios of 10 and less, the dynamic design stress, f ds , can be computed per Equation 4.7 as the product of the minimum specified yield stress, f y ,
a ratio of the expected yield stress to the minimum specified yield stress, R y , and
a Dynamic Increase Factor, c, namely,
f ds = f dy = f y R y c
(4.7)
where R y can be taken as either (a) 1.1 for f y of 50 ksi (345 MPa) or less and 1.0
otherwise, or (b) per Table I-4-1 of the AISC Seismic Provisions for Steel Buildings (American Institute of Steel Construction 2005) for new structural steel, c is
given in Table 4.1 or established on the basis of test data, and f dy is the dynamic
yield stress. For evaluation of older structural members constructed with A36
steel, R y should be set equal to 1.1 unless test data indicate that a greater value
is appropriate.
For ductility ratios of greater than 10, the UFC acknowledges the increase in
steel stress due to strain hardening per Equation 4.8. The relationship between
f ds and ductility ratio must be considered approximate because displacements
and peak strains do not scale linearly. Importantly, the technical basis for the
increase in stress beyond the dynamic yield stress is not provided in the UFC.
f ds = f dy +
f du − f dy
4
(4.8)
where f dy is the dynamic yield stress per Equation 4.7, and f du is the dynamic
ultimate stress. In the UFC, the dynamic ultimate stress is taken as the product
STRUCTURAL STEEL
127
Figure 4.5 Steel Cross Section Distorted Due to Air-Blast Loading (Courtesy of
American Institute of Steel Construction)
of the specified minimum tensile strength and c per Table 4.1. Alternately, the
dynamic design stress at high deformation can also account for the expected
increase in tensile strength above the specified minimum value, namely,
f ds = f dy +
c Rt f y − f dy
α
(4.9)
where values of R y are provided in Table I-4-1 of Seismic Provisions for Steel
Buildings (American Institute of Steel Construction 2005), and α is specified by
the user, but can be taken as 4.
4.2.5 Failure Modes of Structural Components
Introduction This section of the handbook presents introductory information
on typical failure modes (flexural response, shear response, combined flexural and axial response, instability) of structural shapes, with emphasis on
W-shapes loaded in the plane of the web. The focus is on design of new structural
components rather than assessment of existing components.
For assessment of existing structural components whose geometries and/or
materials do not comply with modern standards such as AISC (American
128
MATERIALS PERFORMANCE
Institute of Steel Construction 2005, 2006), the designer is encouraged to
use first-principles mechanics to determine whether performance is acceptable.
Knowledge of the component deformation demands may enable a relaxation of
the element compactness and component bracing rules required by Seismic Provisions for Structural Steel Buildings (AISC 2005) and adopted by-and-large
below.
Flexure Procedures for the computation of flexural strength of prismatic cross
sections as a function of unbraced length of the compression flange and element
compactness are well established and reported in textbooks and the literature. For
the simplest case, where the unbraced length is zero, the cross section is compact,
and the strain rate is low, the flexural strength is computed as the product of the
plastic section modulus and the minimum specified yield stress.
For component checking involving blast load effects, the flexural strength can
be calculated as the product of the plastic section modulus and the dynamic design stress per Section 4.2.3, provided the cross section is seismically compact
per Table I-8-1 of AISC (2005) and laterally braced per Section 9.8 of AISC
(2005). Lateral bracing of both flanges is generally required to address rebound.
Shear The capacity of a steel beam or column in shear can be computed as the
product of the shear area and the dynamic yield stress in shear. For a W-shape
loaded in the plane of its web, the shear area shall not exceed the web area. The
dynamic yield stress in shear, f dv , may be computed as
f dv = 0.60 f dy
(4.10)
where f dy is calculated per Equation 4.7. At the connection of such a W-shape
beam to the flange of a W-shape column, the distributions of stress and strain
will not follow beam theory, and a significant percentage of the shear force in
the beam will be transmitted to the column via the beam flanges (e.g., Kim et al.
2002).
Flexure and Axial Load For checking components subjected to blast load effects that produce flexure and axial load, the procedures of Section 9 of AISC
(2005) and AISC (2006) should be followed to compute coexisting axial and
flexural strengths.
Splices in components subjected to both flexure and axial load should develop
the capacity of the cross section computed assuming the dynamic design stress.
Flexure-axial-shear interaction need not be considered for a W-shape section
loaded in the plane of its web if the blast-induced shear force is less than shear
capacity of the cross section calculated using Equation 4.10 and the ratio of the
plastic section modulus of the web alone to that of the W-shape is small.
Element and Component Instability Element instability (e.g., flange and
web local buckling, web crippling and tearing) and/or component instability
REINFORCED CONCRETE
129
(e.g., lateral-torsional buckling) can prevent a cross section from attaining its
plastic capacity.
Design and detailing of new structural components to resist blast load effects should follow the AISC Seismic Provisions for Structural Steel Buildings
(AISC 2005) unless the component deformations are small. For component ductility demands of 3 and greater, cross sections should be seismically compact per
Table I-8-1 of the AISC Seismic Provisions. For component ductility demands
of less than 3, cross sections should be compact per Table B4.1of the AISC Steel
Construction Manual (AISC 2006).
Crippling or tearing of webs in W-shape cross sections may result from the application of concentrated loads by other framing members. Design against crippling or tearing should conform with Section J10 of AISC (2006), with no allowance for increase in yield stress above the expected value due to strain rate
because the rise time of the loading will likely be much longer than the duration
of the air-blast loading.
Lateral bracing of W-shape flanges to delay lateral-torsional buckling should
generally follow the rules of AISC (2005) for seismic design. Alternate limits
on unbraced length that vary as a function of the component deformation are
provided in UFC 3-340-02 and can be used in lieu of those presented in Section
9.8 of AISC (2005).
4.3 REINFORCED CONCRETE
4.3.1 Stress-Strain Relationships for Concrete
Many textbooks present stress-strain relationships for concrete in uniaxial compression. Sample relationships are presented in Figure 4.6 from Wight and
MacGregor (2008) for concrete with compressive strength ( f c ) ranging between
4500 psi (31 MPa) and 17500 psi (120 MPa). As with structural steel, the stressstrain relationships are established using an ASTM-standard test, which is conducted at a low stress rate of the order of 30 psi per second and causes failure
in 90 to 180 seconds. An increase in compressive strength is accompanied by
increase in the modulus of elasticity (secant modulus in the figure to a stress of
approximately0.5 f c ), an increase in the strain at maximum stress, an increase in
the slope of the softening branch of the stress-strain relationship, and a decrease
in the maximum concrete strain. The modulus of elasticity in units of psi for
normal-weight concrete with a density of 145 lb/ft3 is often taken per ACI 318
(American Concrete Institute 2008) as
E c = 57000 f c
(4.11)
where f c is in units of psi. Values of compressive strength are used to compute
mechanical properties for design after modification for strain-rate effects, as described in Section 4.3.5 below.
130
MATERIALS PERFORMANCE
Figure 4.6 Stress-Strain Relationships for Concrete in Uniaxial Compression at Low
Speed (From Wight and MacGregor 2008, with permission)
Wight and MacGregor present analytical approximations to the uniaxial
stress-strain relationship for concrete, which are not repeated here, and a stressstrain relationship for normal-weight concrete in tension, which is reproduced in
Figure 4.7. The tensile strength of concrete, f t , ranges between 8 and 15 percent
of the compressive strength, with the value dependant on the type of test used for
the measurement (Wight and MacGregor 2008).
The stress-strain relationship for plain concrete is a function of strain rate.
Bischoff and Perry (1991), Ross et al. (1995), Malvar and Ross (1998), Hentz
et al. (2004), Schuler et al. (2006), and Hao and Zhou (2007), among many others, have reported on the effect of strain rate on either compressive or tensile
strength. Sample results are presented in Figure 4.8 (Bischoff and Perry 1991)
for compressive strength and Figure 4.9 (Schuler et al. 2006) for tensile strength.
The vertical axis in each figure is the ratio of dynamic stress to static stress.
REINFORCED CONCRETE
131
Figure 4.7 Generic Stress-Strain Relationship for Concrete in Tension (From Wight
and MacGregor 2008, with permission)
Figure 4.8 Effect of Strain Rate on Compressive Strength (From Bischoff and Perry
1991, with permission)
132
MATERIALS PERFORMANCE
Figure 4.9 Effect of Strain Rate on Tensile Strength of Concrete (From Schuler et al.
2006, with permission)
Appreciable increases in both compressive and tensile strength are evident at
strain rates greater than 0.1 s−1 .
4.3.2 Stress-strain Relationships for Reinforcement
Typical stress-strain relationships for steel reinforcement (or rebar) are presented
in standards and textbooks. Figure 4.10 presents a typical stress-strain relationship for Grade 60 rebar from Malvar (1998). Grade 60 reinforcement is produced in accordance with a number of ASTM standards, including ASTM A615
and ASTM A704. ASTM A706 rebar is often used for seismic and blast applications where weldability, ductility, and bendability are important (Wight and
MacGregor 2008).
As with structural steel, the stress-strain relationship for steel reinforcement
is a function of strain rate. Sample data from Malvar (1998) showing the effect
of strain rate on yield stress and tensile strength are presented in Figure 4.11 and
4.12. The greatest increases in yield stress and tensile strength are observed for
the lower grades of rebar, and the percentage increases in yield stress are greater
than those in tensile strength.
4.3.3 Constitutive Modeling of Concrete and Rebar
Introduction Finite element models of reinforced concrete components and
structures for air-blast analysis will generally use either solid elements to mesh a
REINFORCED CONCRETE
133
Figure 4.10 Typical Stress-Strain Relationship for Grade 60 Steel Reinforcement
(From Malvar 1998, with permission)
cross section along a span or a beam-column element. If solid elements are used
to discretize a cross section, appropriate constitutive models must be selected
for the concrete and reinforcement, and algorithms and elements introduced to
model bond between the concrete and reinforcement. The following subsections
of Section 4.3.3 introduce constitutive models for plain concrete and reinforcement. If beam-column elements are used to model and analyze a reinforced
Figure 4.11 Effect of Strain Rate on Yield Stress of Steel Reinforcement (From Malvar
1998, with permission)
134
MATERIALS PERFORMANCE
Figure 4.12 Effect of Strain Rate on Tensile (Ultimate) Strength of Steel Reinforcement (From Malvar 1998, with permission)
concrete subjected to blast loading, the effects of strain rate (and temperature
if necessary) must be accommodated indirectly, as described in Section 4.3.4.
Plain Concrete Many material models for plain concrete at low strain rate have
been developed. Hao and Zhou (2007) note that these concrete material models
are often modified for high strain-rate loadings by multiplying important low
strain-rate properties such as compressive and tensile strength by a dynamic increase factor that is strain-rate-dependent. The interested reader is directed to
the literature, including Century Dynamics (2005), LSTC (Livermore Software
Technology Corporation 2003), Hao and Zhou (2007), and Tu and Lu (2009),
for information on these models, their implementation in hydrocodes and finite
element codes, and detailed bibliographies. Only increases in compressive and
tensile strength due to strain-rate effects are summarized below.
Recommendations for the compressive strength and/or tensile strength of concrete as a function of strain rate can be found in Soroushian et al. (1986), CEB
(Comit´e Euro-International du B´eton 1993), Malvar and Ross (1998), and Hao
and Zhou (2007), among others.
For compressive strength, CEB (1993) writes that the ratio of dynamic to
static strength (C(˙ε)), is
1.026α
ε˙
f cd
=
C(˙ε) =
f cs
ε˙ s
0.33
ε˙
=γ
ε˙ s
for ε˙ ≤ 30 s−1
(4.12)
for ε˙ > 30 s−1
REINFORCED CONCRETE
135
where f cd is the dynamic compressive strength at strain rate ε˙ in the range of
30 × 10−6 s−1 to 300 s−1 , f cs is the static compressive strength at a reference
strain rate ε˙ s of 30 × 10−6 s−1 , log γ = 6.156α − 2, α = 1/(5 + 9 f cs / f co ), and
f co = 1450 psi (=10 MPa). Malvar and Crawford (1998) noted that the CEB
model for compressive strength properly fitted the data available at the time of
their writing.
Gebbeken and Rupert (2000) proposed the multiplier of Equation 4.13 for
strength increases in compression (C) and tension (T),
Fm
− 1 Wy
C(η) = T (η) = (1 + tanh[0.4(log η − 2)]) ×
Wy
(4.13)
where η is a strain rate normalized by a reference value of 1 s−1 , log is the common logarithm, Fm is a variable that limits the increase at very high strain rates,
and W y is a variable that controls the shape of the function. Values of the variables are different in tension and compression (Gebbeken 2009).
Johnson and Holmquist (1992) presented the multiplier of Equation 4.14 for
strain-rate-related strength increases in tension and compression,
C(η) = T (η) = 1 + cLn(η)
(4.14)
where c is a constant equal to 0.007 (Hao and Zhou 2007), Ln is the natural
logarithm, and all terms have been defined previously. An identical equation
was adopted by Riedel et al. (1999) for the RHT model and implemented in
AUTODYN (Century Dynamics 2005).
For tensile strength, CEB (Comit´e Euro-International du B´eton 1990) writes
that the ratio of dynamic to static strength (T (˙ε )), is
1.016δ
ε˙
f td
=
T (˙ε ) =
f ts
ε˙ s
0.33
ε˙
=β
ε˙ s
for ε˙ ≤ 30 s−1
(4.15)
for ε˙ > 30 s
−1
where f td is the dynamic tensile strength at strain rate ε˙ in the range of 3 ×
10−6 s−1 to 300 s−1 , f ts is the static tensile strength at a reference strain rate ε˙ s of
3 × 10−6 s−1 , log β = 7.112δ − 2.33, δ = 1/(10 + 6 f cs / f co ), and f co = 10 MPa
(=1450 psi). Malvar and Ross (1998) modified the CEB equations as follows,
δ
f td
ε˙
=
for ε˙ ≤ 1 s−1
f ts
ε˙ s
0.33
ε˙
=β
for ε˙ > 1 s−1
ε˙ s
T (˙ε ) =
(4.16)
136
MATERIALS PERFORMANCE
where f td is the dynamic tensile strength at strain rate ε˙ in the range of 10−6 s−1
to 160 s−1 , f ts is the static tensile strength at a reference strain rate ε˙ s of 10−6 s−1 ,
log β = 6δ − 2, δ = 1/(1 + 8 f cs / f co ), and f co = 10 MPa (=1450 psi). Hao and
Zhou (2007) proposed an alternate formulation based on the data of Figure 4.9,
namely,
T (˙ε) = 1.0
= 2.06 + 0.26 log(˙ε)
= 2.06 + 2.0 log(˙ε)
for ε˙ ≤ 1 s−1
for 10−4 s−1 ≤ ε˙ ≤ 1 s−1
for 1 s−1 < ε˙ ≤ 103 s−1
(4.17)
where all terms are defined previously.
Rebar Materials models for rebar are generally identical to those for structural
steels as described in Section 4.2.2: Zhou et al. (2008) used the Johnson and
Cook (1983) material model to characterize the high-strain-rate response of rebar
in slabs constructed of conventional concrete and steel-fiber-reinforced concrete.
4.3.4 Component Level Strain-Rate Effects
Similar to steel structures, most reinforced concrete components will be analyzed and designed for air-blast effects using nonlinear single-degree-of-freedom
(SDOF) models and an indirect treatment of strain-rate effects, as introduced in
Section 4.2.3.
UFC 3-340-02 (U.S. Department of Defense 2008) provides guidance on the
inclusion of strain-rate effects for the computation of component axial, flexural,
and shear strength. The tensile strength of concrete is ignored for axial and flexural strength calculations. Strain-rate effects are addressed using stress-increase or
dynamic increase factors for material strength. Information is provided for concrete compressive and tensile strength and steel reinforcement yield stress and
tensile (ultimate) strength in alternate formats, including functions of strain-rate
and either far-design (low pressure) or close-in design (high-pressure) loadings.
As with calculations for structural steel, the UFC presents simple calculations
for average strain rate for concrete and rebar, which is assumed to be constant.
The calculations are based on an estimate of the time to yield of reinforcement
and the strain in the materials at yield of the cross section. For rebar, the average
strain rate is computed as
ε˙ s =
f dy
Ete
(4.18)
where ε˙ s is the average strain rate in the elastic range of response, E is the elastic
modulus for steel rebar (29,000 ksi, 200 GPa), te is the time to yield of the cross
section (generally computed using response charts), and f dy is the dynamic yield
stress, which is a function of minimum specified yield stress, a stress (dynamic)
REINFORCED CONCRETE
137
increase factor, and a factor to relate expected strength to minimum specified
strength that varies as a function of rebar grade. The variable f ds is the rebar
equivalent of the dynamic design stress for structural steel that was introduced
in Section 4.2.4. For concrete, the average strain rate, ε˙ c , is computed as
ε˙ c =
εc0
te
(4.19)
where εc0 is the strain at maximum concrete stress, which can be taken as 0.002
(see Figure 4.6).
Calculations of strain rate made using Equations 4.18 and 4.19 should be used
with care, in the knowledge that they are at best approximate. Consider a singly
reinforced concrete beam for which the neutral axis depth at yield of the cross
section is approximately 0.30 times the effective depth of the tension rebar, and
the yield strain in the rebar after accounting for strength in excess of the nominal
yield stress and rate effects is 0.0025: the maximum concrete strain at yield of
the tension rebar is approximately 0.0011, and the concrete strain at the approximate location of the resultant compressive force is approximately 0.0007, and
substantially smaller than the 0.002 recommended by the UFC. Alternate values
of strain rate for rebar and concrete would be calculated for doubly reinforced
cross sections (with differing values of strain rate in the tension and compression
rebar), columns subjected to blast loads producing flexure and compression or
tension, walls, and slabs.
Table 4.2, which is adapted from UFC 3-340-02, presents Dynamic Increase
Factors (DIF) for rebar (yield stress and tensile strength) and concrete in compression for flexure (bending), axial compression, shear (diagonal tension failure), shear (direct tension), and bond. The UFC writes that the values given in the
table for flexure assume a strain rate of 0.10 s−1 and 0.30 s−1 for both rebar and
concrete in the low-pressure (far-field) and high-pressure (near-field) ranges, respectively; for reinforced concrete members in compression (i.e., columns), the
Table 4.2 Dynamic Increase Factors for Reinforced Concrete (Adapted from DoD
2008)
Low Pressure
High Pressure
Rebar
Action
Yield
Rebar
Tensile
Concrete
Yield
Tensile
Concrete
Flexure
1.17
1.05
1.19
1.23
1.05
1.25
Compression
1.10
1.00
1.12
1.13
1.00
1.16
Shear—DT
1.00
1.00
1.00
1.10
1.00
1.00
Shear—Direct
1.10
1.00
1.10
1.10
1.00
1.10
Bond
1.17
1.05
1.00
1.23
1.05
1.00
138
MATERIALS PERFORMANCE
corresponding strain rates are 0.02 s−1 and 0.05 s−1 , respectively. These default
strain rates are identical to those assumed for structural steel and reported in the
footnotes to Table 4.1.
As with components of structural steel, the UFC recommends greater increases for rate effects for flexure than compression, noting that “slabs, beams
and girders filter the dynamic effects of the blast load . . . . Dynamic load reaching
columns . . . [has] . . . a long rise time of load . . . [which] results in lower strain
rate.” Further, the UFC writes that the values for shear and bond are smaller than
those for bending or compression for the purpose of “the need to prevent brittle
shear and bond failure and to account for uncertainties in the design process for
shear and bond.”
It is not clear whether the use of three significant figures in Table 4.2 is warranted. Given the range of strain rate associated with the categories of low and
high pressure, the use of two significant figures is likely appropriate. For lower
levels of protection, reinforced concrete components are permitted to sustain
maximum deformations associated with the tensile strength of the rebar. However, component velocity and thus strain rate at the stage of maximum deformation are zero, and material dynamic increase factors should be set to unity.
If a reinforced concrete component is subjected to combined axial force and
bending moment (e.g., a column subjected to bending due to direct air-blast loading on the column flange and axial force due to indirect loading through air-blastinduced beam/girder reactions), one value of the dynamic increase factor should
be used for component strength calculations and deformation calculations, unless the component yield surface can be updated step-by-step in the analysis.
The chosen value of the factor should reflect the dominant contributor to the
combined strength ratio in the beam-column.
4.3.5 Mechanical Properties for Design
The flexural, axial, and shear strength of reinforced concrete components are
typically computed using cross section properties and rebar stresses and concrete strengths. For blast-resistant design in accordance with the Standard, these
stresses and strengths can be increased to a dynamic design stress by factors to
account for strain-rate effects (see Section 4.3.3) and material strengths in excess
of minimum specified values.
UFC 3-340-02 presents procedures for computing dynamic design stresses
and strengths. As for structural steel, the stresses and strengths are tied to deformations but support rotation computed using SDOF analysis is used instead of
ductility ratio (see Section 4.2.4). Table 4.3 presents the UFC recommendations
for dynamic design stresses and strengths for the purpose of design, organized by
action: flexure, shear (diagonal tension and direct), and compression. For flexure and support rotations, θm , of less than 2 degrees, compression rebar can be
ignored for the calculation of flexural strength; for support rotations in excess of
2 degrees, the concrete is assumed to have crushed, and the flexural strength is
provided by a couple between the tension and compression rebar. An increase
REINFORCED CONCRETE
Table 4.3
2008)
Dynamic Design Stresses for Reinforced Concrete (Adapted from DoD
Action
Rebar Type
Flexure
Tension and
Compression
Support
Rotation,
Degrees
Rebar, f ds
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy 1
f dy + ( f du − f dy )/4
( f du + f dy )/2
Dynamic Design Stress
Shear—DT3
Stirrups
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
f dy
f dy
Shear—DT
Lacing
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
f dy + ( f du − f dy )/4
( f du + f dy )/2
Shear—Direct
Diagonal
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
( f du + f dy )/2
f dy
f dy
Compression
139
All
5
Concrete, f dc
f dc
2
2
f dc
f dc
f dc
f dc
f dc
f dc
f dc
4
4
f dc
1
For tension rebar only.
Concrete assumed crushed and moment resistance provided by tension-compression rebar couple.
3
Diagonal tension.
4
Concrete considered ineffective and shear resisted by diagonal rebar only.
5
Component capacity not a function of support rotation.
2
in support rotation corresponds to an increase in rebar strain (and stress), as
indicated in the table. The equation used to compute dynamic design stress for
support rotations 2 < θm ≤ 6 degrees is identical to Equation 4.8 for structural
steel at large deformations.
Dynamic design stresses (strength) for rebar are computed in a similar manner to that described in Section 4.2.4 for structural steel. Equation 4.7 is used
to compute the dynamic yield stress. For ASTM A615 Grade 60 and A706 rebar, the UFC recommends R y equal to 1.1, which results in an expected yield
stress of 66 ksi (455 MPa). Again, as with structural steel, the dynamic ultimate
stress for rebar is taken as the product of the specified minimum tensile strength
and a dynamic increase factor. For concrete, the dynamic compressive strength
is taken as the product of the specified (static) compressive strength and a dynamic increase factor; no allowance is made for concrete strength in excess of
the minimum specified value.
The dynamic increase factors for steel rebar and concrete can be established
using the values provided in Table 4.2, using the data presented in Sections 4.3.1
and 4.3.2, or using charts presented in the UFC, two of which are reproduced
below for concrete compressive strength (Figure 4.13) and rebar yield stress
(Figure 4.14). The curves provided in these figures express dynamic increase
140
MATERIALS PERFORMANCE
Figure 4.13 Effect of strain rate on compressive strength of concrete with f c in the
range of 2500 to 5000 psi (From DoD 2008)
Figure 4.14 Effect of strain rate on the yield stress and tensile strength of rebar (From
DoD 2008)
REINFORCED CONCRETE
141
factors as a function of strain rate. Figure 4.13 represents a reasonable lower
bound on the data of Bischoff and Perry and the CEB recommendations of Equation 4.12 for a concrete strength of 50 MPa (7250 psi) and strain rates of 10 s−1
and less. Note that the strain rate in this figure is presented in units of msec.
The curves of Figure 4.14 for the yield and tensile strength of rebar are in close
agreement with the recommendations of Malvar (1998).
4.3.6 Component-Level Failure Modes
Introduction This section of the handbook presents introductory information
on typical failure modes (flexural response, shear response, combined flexural
and axial response) of reinforced concrete members. The focus is on design of
new structural components rather than assessment of existing components. The
reader is referred to the literature for information on external reinforcement of
concrete elements for blast resistance using steel plate and carbon fiber polymeric materials (e.g., Crawford et al. 2001, Buchan and Chen 2007).
Exactly as with structural steel, the designer is encouraged to use firstprinciples mechanics to determine whether performance is acceptable. The failure mechanisms considered here assume that the core of the reinforced concrete cross section is intact. Spalling and scabbing may occur, and the depth
of reinforced sections should be restricted to core dimensions in such cases.
Reinforced concrete components are susceptible to brisance, or shattering, for
near-field detonations. The likelihood of shattering should be checked, and the
reader is directed to the literature for appropriate procedures.
The following subsections address the common failure modes of reinforced
concrete components under blast loading. Much additional information on
failure modes can be found in the literature, including UFC 3-340-02 (U.S.
Department of Defense 2008), Krauthammer’s Modern Protective Structures
(Krauthammer 2008), Mays and Smith’s Blast Effects on Buildings (Mays and
Smith 1995), and Smith and Hetherington’s Blast and Ballistic Loading of Structures (Smith and Hetherington 1994).
Flexure Procedures for the computation of flexural strength of prismatic cross
sections are well established. Flexural failure can occur in either brittle or ductile modes, depending upon the volume of steel reinforcement in the flexural
element. Economical design of reinforced concrete components to resist blast
loads generally requires ductile detailing to accommodate the large expected deformations while maintaining near-peak strength. Confinement in the form of
seismic hoops or lacing may be required.
For component checking involving blast effects, the flexural strength can be
computed using the dynamic design stresses of Table 4.3 and compressiontension force couples. If the support rotations are less than 2 degrees, flexural
strength is computed using traditional procedures that assume that the tension
rebar yields prior to the concrete in the compression block crushing. For support rotations in the range of 2 to 6 degrees, the cover concrete is assumed to be
142
MATERIALS PERFORMANCE
lost, equal amounts of compression and tension rebar should be provided, and
the flexural strength is computed using a force-couple involving the compression and tension rebar. For support rotations in excess of 6 degrees and less than
12 degrees, equal amounts of compression and tension rebar must be provided,
the flexural strength is computed as a force-couple between the compression and
tension rebar, and the rebar layers must be adequately tied. Shear reinforcement
in the form of ties (stirrups) or lacing is generally required to sustain support
rotations greater than 2 degrees. UFC 3-340-02 provides guidance on the design
of such reinforcement.
For robustness, the minimum longitudinal reinforcement in slabs and beams
must meet the requirements of ACI 318 for load combinations not involving
blast effects. UFC 3-340-02 also requires a minimum amount of tension and
compression flexural rebar to be provided in slabs, but not less than that required
to resist the effects of direct air-blast loading and rebound. These limits for slabs
tend to be smaller than those of Section 10.5 of ACI 318, although a minimum
amount of compression rebar must be provided in all cases.
Shear Four modes of shear failure in reinforced concrete components must be
addressed: (1) diagonal tension, (2) diagonal compression, (3) direct shear, and
(4) punching shear. Critical cross sections are at a distance (a) equal to the effective depth of the cross section from the support for diagonal tension and compression checks for loadings that produce compression in the support and at the
face of the support otherwise, and (b) at the face of the support for direct shear.
Punching shear is checked for slabs around isolated supports (e.g., columns) and
applied loads. Concrete and rebar strengths for component checking can be increased for strain-rate effects per Tables 4.2 and 4.3.
Design against failure by diagonal compression or tension follows the procedure set forth in Chapter 11 of ACI 318. Diagonal compression failure is avoided
in beams (slabs) equipped
stirrups by limiting the nominal shear stress at
with
, where f dc
is the concrete dynamic design stress
the critical section to 10 f dc
per Tables 4.2 and 4.3. Diagonal tension failure is avoided by ensuring that the
sum of the shear resistance provided by the concrete and the stirrups exceeds applied shear force. ACI 318 Sections 11.2.1.1, 11.2.1.2, and 11.2.2.3.can be used
to compute the shear resistance provided by concrete under no axial load, axial compression, and axial tension, respectively. ACI 318 Sections 11.4.7.2 and
11.4.7.4 present equations for establishing the required area of shear reinforcement perpendicular to the axis of the member (i.e., stirrups) and inclined to the
axis of the member (i.e., lacing bars), respectively. The effective depth of the section used in these calculations of concrete and rebar shear resistance is a function
of the expected support rotation and the likelihood of loss of cover concrete. Section 4-18.2 of UFC 3-3340-02 provides appropriate guidance for the calculation
of effective depth.
UFC 3-340-02 requires that shear reinforcement be provided in slabs that
are required to develop large deflections (associated with support rotations
greater than 2 degrees). Shear reinforcement must always be provided to resist
REINFORCED CONCRETE
143
shear stresses in excess of the shear capacity of the concrete. Section 4.18.4
of the UFC provides appropriate guidance for minimum shear reinforcement
in slabs.
Direct shear failure has been observed in blast testing of reinforced concrete
slabs. Failure occurs along a vertical plane at load or stiffness discontinuities
such as the intersection of a slab and its supporting beam or wall. Krauthammer
(2008) provides a discussion of direct shear failure and presents a model to predict the relationship between imposed shear force and deformation (slip across
a vertical plane). Section 4.19 of UFC 3-340-02 presents design equations for
direct shear capacity that includes contributions from concrete and diagonal (inclined) bars. (Stirrups that are placed perpendicular to the plane of a slab provide
no resistance to direct shear because the failure plane is vertical.) Concrete is
assumed to provide resistance to direct shear per Equation 4.20 if the support
rotations are less than 2 degrees (i.e., there is little to no damage to the cover
concrete) and the slab is not subjected to net tension:
bd
Vc = 0.16 f dc
(4.20)
where b is the slab width and d is the effective depth of the cross section. For
support rotations greater than 2 degrees, the direct shear must be resisted by
diagonal (inclined) reinforcing bars only, where the required rebar area per width
b is
Ad =
(Vu − Vc )
f ds sin α
(4.21)
where Ad is the area of inclined rebar per width b, α is the angle between the
inclined rebar and the plane of the longitudinal rebar, Vu is the direct shear force
per width b, Vc is the concrete resistance width b, and f ds is the dynamic design stress for the inclined rebar per Table 4.3. The UFC sets no upper bound
on the applied direct shear stress, likely because the direct shear force is re
. Further, the
sisted by inclined bars only for shear stress in excess of 0.16 f dc
UFC equations do not recognize the benefit of axial compression on the cross
section, which will be present in some components (e.g., columns) and not in
others (e.g., slabs). Upper bounds on direct shear stress are identified in the literature with sample values of 0.35 f c (Ross and Krawinkler 1985) for static loads
(Crawford et al. 2001) for blast loadings. Normal compressive force
and 0.6 f dc
could be included in the formulation using an approach similar to the shearfriction design method of Section 11.6 of ACI 318 or as proposed by Crawford
et al. (2001).
Punching shear must be checked for slabs supported on isolated supports
such as columns. The procedures of Chapter 11 of ACI 318 should be used
to design components to avoid punching shear failure. Section 4.20.1 of
UFC 3-340-02 provides appropriate guidance for the calculation of effective
depth for the punching shear check.
144
MATERIALS PERFORMANCE
Flexure and Axial Load Reinforced concrete components subjected to flexure
and axial load can fail due to either excessive concrete strain (a compressiondominated component) or rebar rupture (flexure-dominated component). The latter can be treated as a flexural element with consideration of axial load. The
procedures of Chapter 10 of ACI 318 should be used to proportion axial-loaddominated components. Appropriate consideration should be given to strain-rate
effects per Section 4.3.4 to compute material stresses for component strength
calculations.
Reinforced concrete components such as slabs may be restrained against
movement in their plane by adjacent framing. Such restraint can give rise
to membrane effects. Krauthammer (2008) notes that compressive membrane
forces at relatively low lateral displacements can enhance the ultimate capacity
of a slab by increasing the flexural strength of the slab at the critical sections.
At large lateral displacements, associated with the loss of core concrete, resistance to blast loadings can be provided by the layers of rebar acting as a cable
net. For such a net to form, the rebar must be appropriately developed into the
supports, and the adjacent framing must be capable of developing the associated
reactions.
Scabbing Scabbing is the loss or separation of cover concrete from the core
concrete associated with large deformations in the component.
Spalling Spalling is similar to scabbing but occurs as a result of the reflection
of a compressive shock wave from the rear face of a concrete component. If the
resulting tensile stresses exceed the tensile strength of the concrete, concrete will
spall. The likelihood of spalling can be computed using codes such as CONWEP
(Hyde 1992).
4.4 STRENGTH-REDUCTION FACTORS FOR STEEL
AND REINFORCED CONCRETE
Strength-reduction factors, denoted φ in most materials standards, are used to
pre-multiply nominal component strength to compute design strength. Design
strengths are generally compared with factored load effects to determine the adequacy of a design. The values assigned to strength-reduction factors range between 0.6 and 0.9, with smaller values assigned to component actions that are
brittle (e.g., shear) and larger values to component actions that are ductile (e.g.,
flexure in the absence of axial force).
There is no industry-wide consensus on the values of φ for blast-resistant
design. Traditional practice has set φ equal to 1.0, and this decision can be
rationalized on the bases that (a) the weapon effect (air-blast) is idealized and
likely conservative, especially for near-field detonations, (b) equivalent SDOF
REFERENCES
145
models are used for load-effect calculations, and (c) strain-rate effects are estimated conservatively.
Blast analysis calculations performed using a finite element code use material
constitutive models to assess stress and strain at the cross section level, deformations at the component level, and displacements at the global level. Best estimate
(mean) material properties are used as input to these constitutive models with no
reduction to design values, that is, φ = 1.
REFERENCES
American Concrete Institute (ACI). 2008. Building Code Requirements for Structural
Concrete (ACI-318-08) and Commentary (ACI 318R-08). Farmington Hills, MI:
American Concrete Institute.
American Institute of Steel Construction (AISC). 2005. Seismic Provisions for Structural Steel Buildings (ANSI/AISC-341s1-05). Chicago, IL: American Institute of Steel
Construction.
. 2006. Steel Construction Manual. 13th Ed. Chicago, IL: American Institute of
Steel Construction.
. 2009. Blast Protection of Buildings (ASCE Standard). Reston, VA: American
Society of Civil Engineers.
Biggs, J. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill.
Bischoff, P. H. and S. H. Perry. 1991. Compressive behavior of concrete at high strain
rates. Materials and Structures 24 (6): 425–450.
Brockenbrough, R. L. and F. S. Merritt. 2006. Structural Steel Designers Handbook. New
York: McGraw-Hill.
Bruneau, M., C. M. Uang, and A. S. Whittaker. 1997. Ductile Design of Steel Structures.
New York: McGraw-Hill.
Buchan, P. A. and J. F. Chen. 2007. Blast resistance of FRP composites and polymer
strengthened concrete and masonry structures—A state-of-the-art review. Composites
Part B: Engineering 38 (5–6): 509–522.
Campbell, J. D. and W. G. Ferguson. 1970. The temperature and strain rate dependence
of the shear strength of mild steel. Philosophical Magazine 21 (169): 63–82.
Century Dynamics. 2005. AUTODYN Users Manual Version 6.1. Horsham, West Sussex,
UK: Century Dynamics.
Chang, K. C. and G. C. Lee. 1987. Strain rate effect on structural steel under cyclic
loading. Journal of Engineering Mechanics 113(9): 1292–1301.
Comit´e Euro-International du B´eton (CEB). 1993. CEB-FIP Model Code 1990. Trowbridge, Wiltshire, UK: Redwood Books.
Crawford, J. E., L. J. Malvar, K. B. Morrill, and J. M. Ferritto. 2001. Composite retrofits
to increase the blast resistance of reinforced concrete buildings (P-01-13). Proceedings
of the 10th International Symposium on Interaction of the Effects of Munitions with
Structures 2001. San Diego, CA.
Gebbeken, N. and M. Ruppert. 2000. A new material model for concrete in high dynamic
hydrocode simulations. Archive of Applied Mechanics 70 (7): 2509–2524.
146
MATERIALS PERFORMANCE
Gebbeken, N. 2009. Personal communication.
Hao, H. and X. Q. Zhou. 2007. Concrete material model for high-rate dynamic analysis. Proceedings of the 7th International Conference on Shock and Impact Loads on
Structures, Beijing, October 2007.
Hentz, S. F., V. Donze, and L. Daudeville. 2004. Discrete element modelling of concrete
submitted to dynamic loading at high strain rates. Computers and Structures 82 (29 30):
2509–2524.
Hyde, D. W. 1992. User’s Guide for Microcomputer Program CONWEP, Application of
TM5-855-1: Fundamental of Protective Design for Conventional Weapons. Instruction
Report SL-88-1. Vicksburg, MS: Structures Laboratory, U.S. Army Waterways Experiment Station.
Johnson, G. R. and W. H. Cook 1983. A constitutive model and data for metals subjected to large strains, high strain rates and high temperatures” Proceedings of the 7th
International Symposium on Ballistics 1983. The Hague, Netherlands.
Johnson, G. R. and T. J. Holmquist. 1992. A computational constitutive model for brittle
materials subjected to large strains, high strain rates and high pressures. In Shock-Wave
and High-Strain-Rate Phenomena in Materials, 1075–1081, edited by M. A. Meyers,
L. E. Murr, and K. P. Staudhammer. New York: Dekker.
Kim, T., A. S. Whittaker, A. J. S. Gilani, V. V. Bertero, and S. M. Takhirov. 2002. Coverplate and flange-plate steel moment-resisting connections. Journal of Structural Engineering 128 (4): 483–491.
Klepaczko, J. R. 1994. An experimental technique for shear testing at high and very
high strain rates. The case of a mild steel. International Journal of Impact Engineering
15 (1): 25–39.
Krauthammer, T. 2008. Modern Protective Structures. Boca Raton, FL: CRC Press.
Lee, W.-S. and C.-Y. Liu. 2004. The effects of temperature and strain rate on the dynamic
flow behavior of different steels. Materials Science and Engineering A 426 (1–2):
101–113.
Livermore Software Technology Corporation (LSTC). 2003. LS-DYNA User’s Manual.
Livermore, CA: Livermore Software Technology Corporation.
Malvar, L. J. 1998. Review of static and dynamic properties of steel reinforcing bars. ACI
Materials Journal 95 (5): 609–616.
Malvar, L. J. and J. E. Crawford. 1998. Dynamic increase factors for concrete. Proceedings of the 28th Department of Defense Explosives Safety Seminar 1998. Orlando,
FL.
Malvar, L. J. and C. A. Ross. 1998. Review of strain rate effects for concrete in tension.
ACI Materials Journal 95 (6): 735–739.
Mays, G. C. and P. D. Smith. 1995. Blast Effects on Buildings. London: Telford.
Meyer, L. W. 1992. Constitutive equations at high strain rates. In Shock-Wave and HighStrain-Rate Phenomena in Materials, 49–68, edited by M. A. Meyers, L. E. Murr, and
K. P. Staudhammer. New York: Dekker.
Meyers, M. A. 1994. Dynamic Behavior of Materials. New York: Wiley.
Moncarz, P. and H. Krawinkler. 1981. Theory and application of experimental model
analysis in earthquake engineering (Report No. 50). Palo Alto, CA: John A. Blume
Earthquake Research Center, Stanford University.
REFERENCES
147
Nicholas, T. and A. M. Rajendran. 1990. Material characterization at high strain rates.
Chapter 3 in High Velocity Impact Dynamics, edited by J. Zukas. New York: WileyInterscience.
Reidel, W., K. Thoma, S. Hiermaier, and E. Schmolinske. 1999. Penetrating of reinforced
concrete by BETA-B-500 numerical analysis using a new macroscopic concrete model
for hydrocodes. Proceedings of the 9th International Symposium on the Interaction of
the Effects of Munitions with Structures 1992, 315–322, Berlin.
Ross, C. A., J. W. Tedesco, and S. T. Kuennen. 1995. Effects of strain rate on concrete
strength. ACI Materials Journal 92 (1): 37–47.
Ross, T. J. and H. Krawinkler. 1985. Impulsive direct shear failure in RC slabs. Journal
of Structural Engineering 111 (8): 1661–1676.
Schuler, H., C. Mayrhofer, and K. Thoma. 2006. Spall experiments for the measurement
of the tensile strength and fracture energy of concrete at high strain rates. International
Journal of Impact Engineering 32 (10): 1635–1650.
Smith, P. D. and J. G. Hetherington. 1994. Blast and Ballistic Loading of Structures.
Oxford: Butterworth-Heinemann.
Soroushian, P., K. B. Choi, and A. Alhamad. 1986. Dynamic constitutive behavior of
concrete. ACI Journal 83 (2): 251–259.
Tu, Z. and Y. Lu. 2009. Evaluation of typical concrete material models used in hydrocodes
for high dynamic response simulations. International Journal of Impact Engineering
36 (1): 132–146.
U.S. Department of Defense (DoD). 2008. Structures to Resist the Effects of Accidental Explosions (United Facilities Criteria UFC 3-340-02). Washington, D.C.: U.S.
Department of Defense.
Wight, J. K. and J. G. MacGregor. 2008. Reinforced Concrete: Mechanics and Design.
5th Ed. Upper Saddle River, New Jersey: Prentice Hall.
Zerilli, F. J. and R. W. Armstrong. 1987. Dislocation-mechanics-based constitutive
relations for material dynamics calculations. Journal of Applied Physics 61 (5):
1816–1825.
Zhou, X. Q., V. A Kuznetsov, H. Hao, and J. Waschl. 2008. Numerical prediction of
concrete slab response to blast loading. International Journal of Impact Engineering
35 (10): 1186–1200.
Zukas, J. A. 2004. Introduction to Hydrocodes. Oxford: Elsevier.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
5
Performance Verification
Curt Betts
5.1 INTRODUCTION
The purpose of this chapter is to familiarize designers who are interested in blast
design with verification of performance for blast and weapons effects and for
stopping vehicles using barriers. The purpose is not to provide guidance on how
to do any of that testing. It is merely to make designers understand when verification and validation should be provided, to make them effective consumers of
performance verification services, and to ensure they understand how to validate
claims made either by other designers or by vendors. This chapter will address
verification by both testing and analysis. It will also address the topic of peer
review.
5.2 PERFORMANCE VERIFICATION
Before the subject of performance verification can be addressed, it is important to
point out that verification has a specific meaning in common use, and it is different from validation. Both involve confirmation, but by convention they confirm
different things in different ways. While their differences in meaning have been
the subject of controversy for decades, the following convention will be used in
this chapter.
Verification refers to the practice of ensuring that a model works for a given
set of conditions. For example, in blast-resistant designs, members may be assumed to be loaded uniformly within a particular scaled range, and they may be
assumed to respond to those loadings in flexure. If a member is closer than that
scaled range, models based on uniform loadings or flexural responses are not
applicable for the member response. Those models cannot be verified for that
application. Verification can also apply to ensuring that a given configuration
falls within the assumptions for meeting a particular performance condition. For
example, for tensile membrane action to be applicable, certain detailing needs
to be verified. Yet another way of applying verification is by checking that all
design requirements have been addressed, that calculations are correct, and that
the design actually reflects what the design analysis says it should.
149
150
PERFORMANCE VERIFICATION
Validation refers to the practice of ensuring that models are technically correct
by comparing them to other validated models or test results. For example, there
are many validated single-degree-of-freedom models for use in blast-resistant
design. Any new model should be validated against previously validated models by running identical member configurations and loadings to ensure that both
codes give similar answers. Alternatively, code results can be compared against
test data to ensure that the models give answers that are consistent with the test
data. Of course, that assumes that the test data are themselves validated or verified, which will lead to the next section of this chapter.
5.3 TESTING
This section will address testing of both building components and vehicle barriers. In both cases there are validated test procedures that can be used to evaluate
performance. Those test procedures are commonly based on some sort of standard test, such as one included in a national standard.
5.3.1 Vehicle Barrier Testing
Vehicle barriers are discussed in Chapter 12, where they are described as antiram structures. Due to the complexities of the structural systems associated with
such barriers, the soil–structure interaction, and the vehicle–structure interaction, designing vehicle barriers solely through analytical means is not sufficient
for validation. Analytical means can be used to design the barriers, but proper
validation requires those models to be validated by vehicle testing.
Any validation testing of vehicle barriers must be based on a standard test
method to be repeatable, to avoid bias, and to ensure that common performance
measures are reported for all vehicle barriers. One such test standard is ASTM
F2656, Vehicle Crash Testing of Perimeter Barriers. It, like any good standard
test, includes specific requirements on how the tests must be set up, how the tests
are run, and how performance is measured. Unlike some standard test methods,
there is a degree of subjectivity to vehicle barrier tests, which is controlled by the
test director who has to decide such things as where the barrier is impacted and
then to determine the validity of the test. Issues such as those are why such standards commonly require testing laboratories to be accredited. Figure 5.1 shows
the end result of a vehicle barrier test.
ASTM F2656, as an example of a test standard, includes details on the vehicles themselves, including gross vehicle weight, age of the vehicles, vehicle
soundness, how ballast needs to be assembled and attached, and in the case of
one of the vehicles, the wheel base. The test standard further establishes requirements for the information to be provided to the test director on the barrier and
information on how the barrier needs to be installed. The standard further establishes detailed criteria for measuring the performance of the barriers, including
the information that has to be collected and evaluated by the test director. That
TESTING
151
Figure 5.1 Vehicle Barrier Test
information includes vehicle acceleration, velocity, penetration, and vehicle and
barrier deformation. It also includes the apparatus necessary to measure those
quantities.
While test standards like ASTM F2656 provide for validation of designs, there
is also an element of verification associated with the barriers. Verification comes
into play when a barrier is specified for installation after testing. Verification at
that point includes reviewing the barrier drawings to ensure that the barrier that
is being installed is the barrier that was tested, that it is installed in a manner that
is consistent with the tested configuration, and that its deployment will result in
its being employed in a manner that is consistent with the way it was tested.
5.3.2 Building Components
Testing of building components similarly requires standard testing procedures to
ensure that tests are run consistently and with standard, comparable performance
measures. Tests can be run at full scale or sub-scale, and using either actual
explosives or blast simulators.
Full-scale testing is the most accurate method to test building components.
It involves building a component to be tested or an entire building to the actual
scale with the threat explosive, and at the applicable standoff distance. Through
152
PERFORMANCE VERIFICATION
Figure 5.2 Full–Scale Building Testing
such tests the full-scale behavior of the building or building component can be
measured or observed to validate compliance with a performance standard.
One common means to set up such a test is to build a building or a test frame
in which the articles to be tested are installed, and to place the threat explosive
at the standoff distance at which the response is to be observed. In the case of
building a whole building, its response and the response of its components are
directly and unambiguously observed. Figure 5.2 shows a full-scale building that
was tested.
In the case of a test frame, the results may need to be evaluated carefully.
Where such a test frame is narrow in comparison to how it will be installed in a
real building, a phenomenon referred to as “clearing” may result in inaccuracies
in the observed component response. Clearing may be an issue when reflected
pressure effects are being measured. With small or narrow targets, the reflected
pressure may be created on the front face of the target, but it may not persist
because the finite boundaries of the target allow the blast wave to propagate
around its edges such that the duration of the load in the target is reduced (the
clearing time), resulting in a reduced load on the component compared to what
it would see were it in a much wider or infinite surface. Clearing is not always
an issue, however. For components such as windows that commonly respond
very quickly to blast waves, the clearing time may not be an issue. Because of
the potential uncertainties in the component response, results from tests where
clearing is an issue need to be evaluated carefully. Figure 5.3 shows test frames.
TESTING
153
Figure 5.3 Test Frames in Arena Configuration
Full-scale testing is often done in what is referred to as an arena test, as shown
in Figure 5.3. In such a test, a number of test objects or test cells or frames
are placed around a test explosive in roughly a circular pattern. Different test
specimens may be at different standoff distances from the explosive, allowing
multiple results from the same test.
Sub-scale testing involves reducing the scale of dimensions associated with
the test, including the test specimen, the distance from the explosive to the test
specimen, and the explosive itself. In those tests, dynamic similitude must be
maintained, which for explosives is done according to what is called Hopkinson
scaling. In Hopkinson scaling, quantities are scaled in proportion to the cube root
of the explosive weight. By scaling, similar blast phenomena can be observed
with smaller explosives at closer distances. For example, refer to Table 5.1 below,
which shows the pressures and impulses for two different explosive weights at
two different standoff distances.
Note that the scaled range (standoff distance / the cube root of the explosive weight) is the same for both tests and that the pressures are the same. That
shows that a test can be run for one explosive weight and can then be scaled upward or downward to reflect the effects of different explosive weights at different
distances. This is a common practice in the blast community. Similarly, other dimensions such as wall heights can be scaled, allowing a wide range of options in
scaled modeling and testing.
It is important to note, however, that not all quantities scale alike. Note that
the impulses associated with the two equivalent pressures are not the same. This
illustrates why scaled testing has to be evaluated and verified carefully. While
such a pair of tests as those in Table 5.1 may be valid for components that are
Table 5.1 Hopkinson Scaled Pressures and Impulses
Quantity
Test #1
Test #2
Weight (W)
1000 lbs TNT
220 lbs TNT
Range (standoff distance)
100 ft
60.4 ft
Scaled range (Range / W1/3 )
10
10
W 1/3
10 lb 1/3
6.04 lb 1/3
Pressure
9.56 psi
9.56 psi
Impulse
80.9 psi-msec
48.7 psi-msec
154
PERFORMANCE VERIFICATION
most responsive to peak pressures, they may not be valid for components that are
more responsive to impulse.
In either full-scale or sub-scale testing, the phenomena that are to be observed
need to be measured carefully and accurately. Most importantly, it is imperative
that the measurements that best reflect the response to be observed are measured.
In all cases the basic geometry of the test needs to be recorded accurately, which
includes the dimensions and construction of the test subject, how it installed into
the test cell or target building, the construction of the test cell or building, the
size and composition of the explosive, and the distance and angle of incidence
from the explosive to the target. In addition, atmospheric conditions such as temperature, humidity, altitude, and wind speed and direction must be recorded.
In measuring test results, there are many measurements that a test report
should include, depending on the nature of the test. Very importantly, the blast
wave must be fully characterized using pressure gages and similar measuring
devices. Then the test specimen response parameters must be measured. Those
measurements depend on the response to be observed. If the member is intended
to remain intact with either elastic or inelastic deformation, the deflection of the
member must be measured. That will allow for calculation of support rotations
or ductility ratios (see Chapter 3). Those are commonly measured using strain
gages or displacement transducers.
If the response is expected to include failure of the test specimen with some
expected hazard level of the resulting fragments, additional measurements would
be required. Those would include measuring the sizes, weights, and translations
of the fragments, and their accelerations. Measuring devices for those include
accelerometers and high-speed digital cameras. Figure 5.4 shows a setup for both
high-speed digital camera recording and deflection. The camera cannot be seen,
but the stripes painted in the wall allow for accurate evaluation of the velocities
of the fragments.
Another option for measuring blast response of components is to use a blast
simulator. The most common blast simulators are tubes into which test specimens are mounted at one end. The blast simulating “driver” is at the end of the
tube opposite the test specimen, and it commonly drives air into the test specimen
at high velocities that simulate a blast wave. The measurements of the responses
of the test specimens are similar to those in open-air tests. Tests in blast simulators can usually be done much less expensively and in much quicker progression
than open-air blast tests, and they can be operated in areas with far fewer restrictions than those in which open-air blasts tests are run. It is important to recognize,
however, that the blast simulator loadings should be validated against open-air
tests to ensure that the loadings are representative of actual conditions. Figure
5.5 shows a shock tube blast simulator with a wall specimen installed into it
There are few test standards for blast testing, but two commonly used standards are ASTM test standards for doors and windows. ASTM F1642, Glazing
and Glazing Systems Subject to Airblast Loadings, is described in Chapter 10.
It describes test setup, test execution, and performance standards for windows
subjected to blast loads or simulated blast loads. ASTM F2247, Standard Test
TESTING
Figure 5.4 Digital Photography and Deflection Measurement
Figure 5.5
Shock Tube Blast Simulator
155
156
PERFORMANCE VERIFICATION
Method for Metal Doors Used in Blast Resistant Applications, provides similar
standards for doors.
It is very important that test results undergo rigid verification, which includes
both validation against other test results and evaluating test reports. The latter is
particularly important. Somebody who understands test reports should carefully
review all of the conditions under which the tests were run, how the measurements were made, and how the test specimen responded. He or she must ensure
that the test was run in a valid manner and that the results are meaningful. In addition to evaluating the conduct and setup of the test, he or she must also verify
that the results are valid based on issues such as accuracy of the loading, whether
scaling was taken into effect correctly, and whether clearing was an issue or was
accounted for. This is less difficult for tests that are conducted in accordance with
a particular standard, but those tests must also be evaluated carefully. In many
ways explosive testing is more of an art than a science.
5.4 ANALYSIS
There are many ways to analyze the response of building elements to blast
shocks, ranging from approximations to highly detailed analyses such as finite
element analysis, and for each of those there are multiple computer codes for
performing that analysis. There are also hand methods and charts. Analysis is
the least expensive means for evaluating performance, but because of the nonlinearity of such analyses and the uncertainties in the loadings, dynamic material
properties, and dynamic structural response, analyses require stringent verification and validation.
Verification includes evaluating the applicability of specific methods to individual situations and ensuring that the models are applied correctly. Depending
on the assumptions upon which models are based, they may have limited applicability. Optimally, the applicability should be verified before they are applied, but
if not, their applicability must be verified in evaluating the results of the analyses.
While the applicability of a model to a given situation needs to be verified,
it is at least as important that models be validated. Models can be validated by
comparing their results to the results of other models that have been validated
for identical situations. For example, a computer code to be validated can be
used to analyze a beam, a wall, a window, or any other building component
of a specific geometry and configuration, and those results can be compared to
those derived by using a previously validated code. This is particularly the case
with finite element analyses, where there are so many variables that have to be
considered in the models that results can appear reasonable, but turn out to be
fallacious.
Results can also be compared to test results, which can serve as benchmarks.
In either case, validation gives the user both confidence in the model and confidence that its application is reasonable for that situation. This is particularly
useful for finite element models.
REFERENCES
157
In addition to analyzing buildings and building components to blast loads,
there are models for analyzing vehicle barriers. Because they model such
complex situations as vehicle–structure interaction, structural response, and
soil–structure interaction, those models tend to be very complicated. They often include finite element analysis models. Because they’re so complicated, it
is especially important that they be validated. Usually that validation is done
by either comparing them to existing test results or performing testing after the
analysis to validate the analysis results. The most common use of analysis models for vehicle barriers, therefore, is to design the barriers and validate the results
through testing.
5.5 PEER REVIEW
Because analysis and testing for blast loads involve such complexity, it is very
important that test results and analyses be validated through peer review. Peer
review is particularly important for structures with unique structural features or
structures that have high importance or high economic value.
In addition, because the number of experienced blast consultants is still limited, peer review is particularly important where analysis is done by less experienced designers.
Verification and validation through peer review can take multiple forms. As a
minimum, it should include detailed review of drawings to ensure that details are
consistent with the assumptions in the design analyses and that the drawingsare
properly detailed, to allow the expected performance to be achieved. Verification
should also evaluate the applicability of the models used and the accuracy with
which they were applied. It should also include checks of at least major elements
of the design calculations.
Peer review should be done by people who have significant experience in
applying the models used for the applications on which they were applied.
REFERENCES
ASTM. 2003. Standard Test Method for Metal Doors Used in Blast Resistant Applications (F2247). West Conshohocken, PA: ASTM International.
. 2004. Glazing and Glazing Systems Subject to Airblast Loadings (F1642).
West Conshohocken, PA: ASTM International.
. 2007. Vehicle Crash Testing of Perimeter Barriers (F2656). West Conshohocken, PA: ASTM International.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
II
Blast Phenomena and Loadings
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
6
Blast Phenomena
Paul F. Mlakar and Darrell Barker
6.1 INTRODUCTION
Blast is a pressure disturbance caused by the sudden release of energy. People
often think of blasts in terms of explosions such as the detonation of an explosive
charge. However, there are many other blast sources that have the potential to
cause damage. For example, chemicals may undergo a rapid decomposition under certain conditions. These events are often referred to as runaway reactions.
Flammable materials mixed with air can form vapor clouds that when ignited
can cause very large blasts. Blasts are not always caused by combustion; they
can also result from any rapid release of energy that creates a blast wave, such as
a bursting pressure vessel from which compressed air expands, or a rapid phase
transition of a liquid to a gas.
The loads resulting from a blast are created by the rapid expansion of the energetic material, creating a pressure disturbance or blast wave radiating away from
the explosion source, as shown in Figure 6.1. Blast pressure is more properly
overpressure, because it is relative to ambient conditions, rather than an absolute
pressure. Shock waves are high-pressure blast waves that travel through air (or
another medium) at a velocity faster than the speed of sound. Shock waves are
characterized by an instantaneous increase in pressure followed by a rapid decay. Pressure waves are lower amplitude and travel below the speed of sound.
Pressure waves are characterized by a more gradual increase in pressure than a
shock wave, with a decay of pressure much slower than a shock wave. In most
cases, shock waves have a greater potential for damage and injury than pressure
waves.
As a blast wave travels away from the source, the pressure amplitude decreases, and the duration of the blast load increases. Overexpansion at the center
of the blast creates a vacuum in the source region and a reversal of gas motion. This negative pressure region expands outward, causing a negative pressure
(below ambient), which trails the positive phase. The negative phase pressure
is generally lower in magnitude (absolute value) but longer in duration than the
positive phase. Generally speaking, positive phase blast loads are more consequential than negative phase loads, the latter of which is often ignored.
161
162
BLAST PHENOMENA
GROUND – REFLECTED WAVE
ASSUMED PLANE
WAVE FRONT
STRUCTURE
GROUND
SURFACE
W
Figure 6.1 Propagating Blast Wave
Expansion of the blast wave causes air particles to move outward during the
positive phase and inward during the negative phase. The flow of air particles
creates a pressure analogous to that caused by wind. The pressure produced by
this flow is referred to as the dynamic pressure. This pressure is lower in magnitude than the shock or pressure wave and imparts a drag load similar to wind
loads on objects in its path.
As the shock or pressure wave strikes a wall or other object, a reflection occurs, increasing the applied pressure on the surface. This reflected pressure is
considerably higher than the incident or free-field pressure wave. At the free
edges of a reflecting surface, the discontinuity between the forward traveling
incident blast wave and rearward traveling reflected blast wave creates a rarefaction, or pressure relief wave. Rarefactions travel inward from the outer edges
across the face of the reflecting surface. The rarefaction waves relieve the positive
reflected pressure down to the free-field or side-on pressure, plus drag pressure.
The peak reflected pressure is not affected, only the duration The time required
for the rarefaction waves to completely relieve the reflected pressure is termed
the clearing time. This clearing time varies across the surface. It should be noted
that a rarefaction wave does not instantaneously clear reflected pressure; rather,
the relief is somewhat gradual and takes longer than the time required for the
leading edge of the rarefaction to travel to the center of the reflected surface. If
the clearing time exceeds the positive phase blast wave duration, clearing does
not affect the positive phase loads.
6.2 SOURCES OF BLASTS
Blasts involving chemical reactions can be classified by their reaction rates
into two primary groups: deflagrations and detonations. A deflagration is an
SOURCES OF BLASTS
163
Pso
PRESSURE
POSITIVE IMPULSE, is
Ps(t)
t A+ t 0
NEGATIVE IMPULSE, is
tA
Po
Pso
tA
POSITIVE
PHASE
DURATION
NEGATIVE
PHASE
DURATION
t0
t 0–
Figure 6.2 Pressure Wave from Deflagration
oxidation reaction that propagates at a rate less than the speed of sound in the unreacted material. The corresponding blast wave is often termed a pressure wave
and has a finite rise time, as illustrated in Figure 6.2. A “fast” deflgration can
create a more sudden rise in pressure. By contrast, in a detonation, the reaction front propagates supersonically, usually many times faster than the speed of
sound. This blast wave is termed a shock wave and has an instantaneous rise in
pressure, as seen in Figure 6.3. Since pressure is closely related to reaction rate,
detonation pressures are usually many times higher than deflagration pressures.
A blast involving an explosive is an exothermic chemical reaction, usually involving an oxidizer and a fuel. Explosion reactions can be produced by a wide
array of materials, some familiar, such as trinitrotoluene (TNT), while others are
less well known. The reaction rate is dependent on the chemical and physical
properties of the energetic material, reactant proportions and homogeneity, geometry of the material, characteristics of the “container” in which the material
resides, method and energy of initiation, and other initial conditions. A deflagration may be initiated by a “soft” ignition source such as friction, spark, or open
flame. Under certain conditions, a deflagration can transition into a detonation.
Alternatively, detonations can be directly initiated in an explosive material that
exceeds certain minimum geometry constraints if it is impinged upon by a shock
source of sufficient strength.
Deflagrations and detonations may involve oxidizers and fuels that are
oxygen-deficient, or materials that may produce flammable gases as a product
of reaction. In either case, the unreacted or flammable products may mix with
air and result in secondary burning. The secondary burning does not contribute
164
BLAST PHENOMENA
Pso
PRESSURE
POSITIVE IMPULSE, is
NEGATIVE
IMPULSE, is–
Ps(t)
AMBIENT
Po
tA
tA+ to+ t –
o
tA+ to
–
Pso
tA
t=0
Ps–(t)
POSITIVE
PHASE
NEGATIVE PHASE
DURATION
to
DURATION
t–
o
TIME AFTER EXPLOSION
Figure 6.3 Shock Wave from Detonation
significantly to the blast pressures for external explosions but can be a major
consideration for predicting internal explosion blast pressures.
Industrial explosive, propellant, and pyrotechnic manufacturing provides a
wide array of energetic material configurations and materials. These products
may include gas generators for airbags, packaged general-purpose explosive
charges, shaped high-explosive charges for cutting, or blasting agents used in
mining. Some materials used in these manufacturing processes may not be well
characterized in terms of potential explosive potential or reactivity hazards.
Therefore the structural analyst should take care to understand the full range
of blast loads that may be produced, and to design and apply appropriate factors
of safety to account for uncertainties.
Terrorist threats principally involve solid materials such as plastic explosives
or improvised explosives such as ammonium nitrate and fuel oil (ANFO). These
materials can readily produce detonations, but the blast strength may be low
due to inefficient configuration or the presence of contaminants. Flammables
can potentially be used, but they are more difficult to transport and successively
initiate.
For blasts originating from a chemical reaction, the material involved and its
state, the proportions of fuel and oxidizer (if applicable), container strength and
configuration, and the method and energy of initiation are among the parameters
that determine blast characteristics. Explosive materials can be broadly categorized based on their state. Solid materials typically produce high-pressure, shortduration loads, while combustible gases of comparable energy produce lowerpressure, longer-duration loads. Energy output and standoff distance in the configuration of interest are key to accurately determining blast loads acting on a
structure. Blast sources are described by group in the following paragraphs.
SOURCES OF BLASTS
165
High explosives are materials that are intended to produce detonation events
with supersonic reaction fronts. The reactions proceeding through the explosive
are self-sustaining if the charge is of sufficient diameter and properly initiated.
Reaction rate, or detonation velocity, varies with material type and is a key factor in detonation pressure for a material. Detonation velocities typically range
from 3000 to 30,000 ft/s (1–9 km/s). High explosives can be classified by their
sensitivity to initiation into primary, secondary, and tertiary explosives. Primary
explosives are very sensitive to initiation—for example, lead azide used in detonators. Secondary high explosives include TNT, composition C4, and dynamite.
Tertiary explosives are much less sensitive to initiation and require a powerful
booster charge to initiate them. The blasting agent ANFO is an example of a
tertiary explosive.
For convenience in predicting blast pressures, the energy release of high explosives is commonly measured as a value relative to that of TNT. This “TNT
equivalence” is used to determine a TNT charge weight capable of producing the
same explosion energy, blast pressure, or blast impulse as the explosive of interest. The TNT equivalence is different for energy, peak pressure, and impulse, and
separate TNT equivalent energy values are reported for many materials. Equivalencies for a number of these materials appear in Table 6.1.
High-explosive detonations may be accidental or intentional events. Accidents
involving high explosives can occur in processing, handling, and transportation.
Intentional detonations can include explosives testing, military weapons, demolition, specialized cutting or explosive forming, and terrorist acts. In the case of
intentional detonations, structures, such as test structures, may be required to
withstand multiple events.
Confined explosive charges will create gas pressures in the structure in addition to shock waves. If the structure is not adequately vented, the effects of these
gas pressure loads may exceed those of the shock loads.
High explosives come in many forms determined by their chemical and physical properties and intended use. Most explosives have additives to aid in stabilizing the material chemically or physically. Powdered explosives may be combined
with a plastic binder to form plastic-bonded or PBX explosive. Extrudable explosives have a viscous form due to the addition of rubber resins to the explosive
material. This addition allows the explosive to be molded or extruded to produce
a particular shape.
Certain liquids, such as nitromethane, can also detonate. Nitromethane is relatively insensitive and must be initiated by a strong ignition source such as a
high-explosive booster. It does have safety advantages over other materials and
may be handled as a simple flammable liquid in many cases. The shock front
produced by nitromethane is well formed and produces blast loads with approximately 100% TNT equivalence.
Propellants and pyrotechnics, also known as low explosives, are energetic materials that do not typically detonate and are used to produce gas, smoke, flash,
or sound. Both solid and liquid rocket propellants, gun propellants, and black
powder are examples of low explosives. Pyrotechnics include many different
166
BLAST PHENOMENA
Table 6.1
TNT Equivalence of High Explosives
Explosive
Density
Mg/m3
Equivalent
Moss for
Pressure
Equivalent
Mass for
Impulse
Pressure
Range MPa
Amatol (50/50)
1.59
0.97
0.87
NA1
Ammonia Dynamite
(50 percent Strength)
1
0.90
2
NA
0.90
NA1
Ammonia Dynamite
(20 percent Strength)
NA1
0.70
0.702
NA1
ANFO (94/6 Ammonium
Nitrate/Fuel Oil)
NA1
0.87
0.872
0.03 to 6.90
AFX-644
1.75
0.732
0.732
NA1
1.59
2
2
NA1
2
AFX-920
1.01
2
1.01
AFX-931
1.61
1.04
1.04
NA1
Composition A-3
1.65
1.09
1.07
0.03 to 0.35
Composition B
1.65
1.11
1.20
0.98
1.30
0.03 to 0.35
0.69 to 6.90
Composition C-3
1.60
1.05
1.09
NA1
Composition C-4
1.59
1.20
1.37
1.19
1.19
0.07 to 1.38
1.38 to 20.70
Cycloid (75/25 RDX/TNT)
(70/30)
(60/40)
1.71
1.73
1.74
1.11
1.14
1.04
1.26
1.09
1.16
NA1
0.03 to 0.35
NA1
DATE
1.80
0.87
0.96
NA1
2
Explosive D
1.72
0.85
0.81
0.01 to 0.30
Gelatin Dynamite
(50 percent Strength)
NA1
0.80
0.802
NA1
Gelatin Dynamite
(20 percent Strength)
NA1
0.70
0.702
NA1
H-6
1.76
1.38
1.15
0.03 to 0.70
HBX-1
1.76
1.17
1.16
0.03 to 0.14
HBX-3
1.85
1.14
0.97
0.03 to 0.17
HMX
NA1
1.25
1.252
NA1
1.80
2
NA1
L.X-14
1
NA
1.80
MINOL II
1.82
1.20
1.11
0.02 to 0.14
Nitrocellulose
1. 65 to 1. 70
0.50
0.502
NA1
Nitroglycerine Dynamic
(50 percent Strength)
NA1
0.90
0.902
NA1
Nitroglycerine (NQ)
1.72
1.00
1.002
NA1
Nitromethane
NA1
1.00
1.002
NA1
(continued)
SOURCES OF BLASTS
167
Table 6.1 (Continued)
Explosive
Density
Mg/m3
Equivalent
Moss for
Pressure
Equivalent
Mass for
Impulse
Pressure
Range MPa
Octol (75/25 HMX/TNT)
(70/30)
1.81
1.14
1.02
1.09
1.06
1.092
NA1
0.01 to 0.30
PBX-9010
1.80
1.29
1.292
0.03 to 0.21
PBX-9404
1.81
1.13
1.70
1.132
1.70
0.03 to 0.69
0.69 to 6.90
PBX-9502
1.89
1.00
1.00
NA1
2
PBXC-129
1.71
1.10
PBXN-4
1.71
0.83
PBXN-107
1.64
PBXN-109
1.67
1
1.05
2
1.05
2
1.10
NA1
0.85
NA1
2
NA1
2
NA1
2
1.05
1.05
PBXW-9
NA
1.30
1.30
NA1
PBXW-125
1.80
1.022
1.022
NA1
Pentolite. (Cast)
1.64
1.68
NA1
1.42
1.38
1.50
1.00
1.14
1.00
0.03 to 0.69
0.03 to 4.14
0.69 to 6.90
PETN
1.77
1.27
1.272
0.03 to 0.69
Picroto1 (52/48 Ex D/TNT)
1.63
0.90
0.93
0.03 to 4.10
RDX
NA1
1.10
1.102
NA1
RDX/Wax (98/2)
1.92
1.16
1.162
NA1
1.30
2
NA1
2
NA
1
TATB
NA
1
1.00
1.00
NA1
Tetryl
1.73
1.07
1.072
0. 02 to 0.14
2
RDX/AL/Wax (74/21/5)
1.30
1.59
1.06
TNETB
1.69
1.13
0.96
0.03 to 0.69
1.75
1.18
123
1.23
1.32
1.38
1.11
1.322
1.382
0.03 to 0.69
NA1
NA1
1.63
1.00
1.00
Standard
TNETB/AL
(90/10)
(78/22)
(65/35)
TNT
1.06
NA1
Tetrytol (75/25 Tetryl/TNT)
Torpex
1.85
1.23
1.28
0.01 to 0.30
Tritonal (80/20 TNT/AL)
1.72
1.07
0.96
0.03 to 0.69
1
2
NA - Data not available.
Value is estimated.
materials such as fireworks, smoke grenades, flares, and the gas generator material in automobile air bags. The relatively low reaction rates of these materials
result in low blast pressure output because they typically don’t detonate. Fire
hazards and gas generation concerns dominate the safety analysis in most cases,
but detonations or deflagrations cannot be ruled out in all cases.
168
BLAST PHENOMENA
Figure 6.4 Vapor Cloud Explosion in 9/11 Pentagon Crash
Predicting blast loads for these materials can be challenging. There is much
less data available for propellants and pyrotechnics than for high explosives, with
respect to blast loads. Burn rates, gas generation, and energy output rates are
available in handbooks for common propellants (Cooper and Kurowsky 1996)
These values can be used to perform a burn simulation and determine the gas
pressure load in a confined area. In some cases, TNT equivalencies for these
materials have been used. However, this is problematic because peak pressure
for propellants and pyrotechnics is much lower than for TNT, but the impulse
of propellents can be higher in a confined space than for the TNT equivalent.
Care must be taken when equivalencies are based on small-scale tests, as large
quantities may react in a more energetic manner.
Vapor cloud explosions involve the release of a flammable material which,
when mixed with air and given the proper conditions, forms an ignitable material, potentially leading to an explosion. The vapor cloud explosion following the
9/11 Pentagon crash is shown in Figure 6.4. Only fuel within flammability limits
participates in the explosion, which is a minor percentage of the total quantity of
fuel released (Baker et al. 1991). Much of the fuel vapor is too lean or too rich to
combust and does not produce blast overpressure. Blast output of a vapor cloud
explosion is highly variable and depends on the fuel, size of the flammable vapor cloud, presence of congestion (obstacles such as structural members, pipes,
and vessels) in the flammable region, and presence of confining surfaces. Release of a flammable material into the open air, absent congestion or confinement, typically results in a fireball that does not produce damaging overpressures.
SOURCES OF BLASTS
169
Confinement and congestion in the flammable cloud create turbulence and accelerate the combustion, thereby increasing the flame speed. Peak overpressure is
a strong function of flame speed; thus, a key element in vapor cloud explosion
prediction is maximum flame speed. Vapor cloud detonations are rare and only
known to occur with high-reactivity materials, strong ignition sources, and/or
severe cases of confinement and congestion. The vast majority of vapor cloud
explosions are deflagrations. A particularly noteworthy characteristic of deflagrations is that high flame speeds only occur in confined/congested regions of
the flammable cloud. Outside of congested areas, flame speed decreases. As a
result, deflagration blast loads are produced in confined/congested areas, rather
than at the point of release or dispersion. In the rare case of vapor cloud detonations, the blast loads are produced by the detonatable material and are typically
considered to be centered in the dispersed cloud and not at the point of release.
Overpressures produced by vapor cloud explosions are substantially lower
than those produced by high explosives. For example, a vapor cloud detonation may reach 16–18 atmospheres overpressure, compared with 100,000 atmospheres or greater for most high explosives. However, the energy of vapor cloud
explosions can be very large, as evidenced by some of the devastating accidents
which have occurred. In Flixborough, U.K., a 30,000-kg release of cyclohexane
caused an explosion that resulted in 28 fatalities, multiple injuries, and destruction of more than 150 structures. The large energy in hydrocarbons produces a
commensurately large impulse. Blast predictions made using a TNT equivalent
approach tend to be inaccurate due to the disparity between the pressure and
impulse compared with high explosives. As a result, TNT equivalence analyses
of vapor cloud explosions are not recommended. A more accurate method is to
perform a dispersion analysis of the flammable release to determine fuel in the
flammability limits, select appropriate speed or reaction severity, then use nondimensional scaled energy curves to determine the pressure and impulse at various standoff distances. Computational fluid dynamics (CFD) analysis is another
method of determining blast pressures for vapor releases.
Vessels designed to contain fluids under pressure have the potential to create
hazardous overpressures and fragments if the vessel fails while under pressure. A
pressure vessel burst may occur at normal working pressure due to a mechanical
integrity problem, weakening, or other external factors (such as impact); or at
elevated pressure due to excessive pressurization from an external source, an
uncontrolled internal chemical reaction, heating vessel contents with insufficient
pressure relief, or other means. This type of accident results in loss of contents
and may lead to a follow-on explosion when the contents subsequently mix with
air, if the contents are flammable or pyrophoric.
In the far field, the blast load from a vessel that fails at high pressures (several
thousand psi [1000 psi = 6895 kPa] or higher) is similar to that caused by high
explosives. In the near field, overpressures are less than those produced by high
explosives, and the use of TNT equivalent prediction procedures will overestimate pressure. Key parameters for estimating blast effects from bursting vessels
170
BLAST PHENOMENA
are the volume of the material in the vessel, the failure pressure, the composition
of the contents, and the temperature of the contents.
Fragments represent an additional hazard from bursting vessels. Failure at
working pressure or overpressurization at a relatively slow rate usually results
in a small number of fairly large fragments. In some cases, one or both ends
of a cylindrical pressure vessel may be propelled long distances by a rocketing
mechanism. Large fragments may be thrown as a result of the expansion of the
vessel contents. Containment of these fragments, which can be quite challenging,
may be required to provide adequate safety. Fragment hazard prediction methods
are discussed further in Chapter 8 of this handbook.
A boiling liquid expanding vapor explosion (BLEVE) occurs when a vessel
containing a liquid stored above its boiling point for atmospheric pressure fails
and a rapid depressurization of the liquid occurs. Upon depressurization, a portion of the superheated liquid rapidly boils, and the expanding vapors contribute
to the blast wave. Compressed gases in the vessel prior to failure also contribute
to the blast wave as in a bursting pressure vessel. A common BLEVE scenario
is a pressure vessel engulfed in an extended fire, causing weakening of the vessel shell to the point of failure. A BLEVE may involve a flammable liquid, and
mixing of the flammable material with air and subsequent ignition may cause a
fireball. A detailed discussion of BLEVE events and of simplified blast prediction methods is provided in CCPS’s Guidelines for Evaluating the Characteristics of Vapor Cloud Explosions, Flash Fires, and BLEVES (Center for Chemical
Process Safety 1994).
6.3 CHARACTERISTICS OF BLAST WAVES
6.3.1 Key Parameters
An idealized pressure-time history indicating the key parameters of a blast load is
shown in Figure 6.3. Figure 6.5 shows a comparison between free-field, or sideon, and reflected pressure-time histories. The parameters shown in these figures
are defined as:
Po
Pso
= Ambient pressure
= Peak positive side-on overpressure
Pso−
Ps (t)
= Peak negative side-on overpressure
= Time varying positive overpressure
Ps− (t) = Time varying negative overpressure
Pr
= Peak reflected overpressure
Is
= Positive-phase-specific impulse, the integration of the positive
phase pressure-time history
−
= Negative-phase-specific impulse, the integration of the negative
is
phase pressure-time history
PRESSURE
CHARACTERISTICS OF BLAST WAVES
171
Pr
Pso
to+ t –
o
to
Po
POSITIVE
NEGATIVE PHASE
PHASE
DURATION, to
DURATION, t o–
TIME
Figure 6.5 Comparison of Free Field and Reflected Blast Loads
ta
to
= Time of arrival
= Positive phase duration
to− = Negative phase duration
The term “overpressure” refers to a gauge pressure, or the blast pressure relative to ambient pressure.
Free-field loads are those produced by blast waves sweeping over surfaces
unimpeded by any objects in their path. This load is also referred to as side-on
when the blast wave sweeps over a wall or other object parallel to its direction
of travel. Side-on pressure terms are indicated by a “so” subscript in the figures
in this chapter, as well as in most references.
6.3.2 Scaling
Blast pressures, load duration, impulse, shock wave velocity, arrival times, and
other blast parameters are frequently presented in scaled form. Research has
172
BLAST PHENOMENA
shown that scaling laws can be applied to explosions, allowing data from one
explosion test to be applied to a geometrically similar (larger or smaller) case.
As a result, scaling has tremendous utility in blast predictions, allowing compilations of explosion test and numerical modeling data to be used to predict loads
for any combination of explosion energy and standoff distance within the range
of data. The most common form of scaling is called “cube root scaling” owing to
the fact that blast parameters are scaled by the cube root of the explosion energy.
Both dimensional and dimensionless scaling are used, and care must be taken
with all unit parameters to ensure the scaling is correctly applied for the blast
curves being used.
6.4 PREDICTION OF BLAST PARAMETERS
This section describes the principal methods for predicting blast parameters, with
some discussion of practical applications and limitations. A more detailed treatment is given for high explosives and vapor cloud explosions because these tend
to be the most common design threats for protective construction. Definition of
the blast hazard is discussed in Chapter 2, and calculation of the loads imposed
on the structure in Chapter 7.
In many cases, blast load prediction is a multi-step process. Often an initial
estimate of blast loads will be made using simplified techniques and a conservative estimate of explosion energy. It may become obvious at this point that
the predicted loads are so low they are not a contributing factor in structural design, or will not cause undue cost or complexity in the design. Thus, the initial
loads are judged to be adequate for analysis or design. If loads are high enough
to drive the design, it may be appropriate to perform a more detailed blast load
prediction which uses less conservative estimates of energy and more detailed
prediction techniques.
The basic simplified approach involves predicting free-field loads using empirical or semi-empirical methods. Blast curves that relate explosion energy,
standoff distance, and blast wave parameters are typically used in this initial
load-prediction step. These curves are readily available for high explosives and
vapor cloud explosions.
6.4.1 High Explosives
A significant amount of data has been produced that quantifies the relationships
among charge weight, standoff distance, and blast parameters for both positive
and negative phases. A Unified Facilities Criteria publication, Structures to Resist the Effects of Accidental Explosions, produced by the Department of Defense, contains the relationships in the form of scaled blast curves (U.S. Department of Defense 2008). This document is publicly available, and an electronic
version also exists to aid in the use of the numerous tables and figures.
PREDICTION OF BLAST PARAMETERS
173
The empirical blast parameter curves provided in this document plot air-blast
parameters versus scaled distance for both air-burst and surface-burst configurations. A charge detonated on the ground will produce higher blast loads than an
air burst due to the reflection of the shock by the ground surface. For unyielding surfaces with the charge located on the surface, the effect is a doubling of
the charge weight, since the energy of the blast directed to the ground is fully
reflected. For soft soils or charges located above the surface, the reflection factor
is less than two, since some of the energy is absorbed by the substrate. Some
guidance for selecting the reflection factor is given in the document, but it is
conservative to assume a fully reflecting surface.
The most commonly used approach to present blast wave relationships for
high explosives is the Hopkinson-Cranz, or cube root, scaling method. Figure 6.6
is a simplified version of TNT blast curves that provides the parameters for
charges located at ground level. The cube root term results from geometric scaling laws in which charge diameter varies in proportion to all distances, and thus
the charge weight is proportional to the cube of the charge diameter. To use these
empirical curves, one computes the scaled distance by dividing the standoff distance from the charge to the point of interest by the cube root of the charge
weight. For explosives, this takes the form of
Z
Z
R
W
= R/W 1/3 where :
= scaled distance (ft/lb1/3 )
= standoff distance (ft)
= explosives weight (lb)
(6.1)
The curves provide pressures, P, which are the same at a given scaled distance. The curves also provide scaled times, t/W 1/3 , and scaled impulses, I/W 1/3 .
The actual times and impulses are then found by multiplying the scaled values
by W 1/3 .
In accordance with the common practice for blast design, the variables in
Equation 6.1 and Figure 6.6 have physical dimensions. However the procedure
is rigorously founded on dimensionless modeling. It is also based on atmospheric conditions at sea level. While this suffices for most practical situations,
the effects of differing ambient conditions can be treated more comprehensively
through Sachs or energy scaling (Baker 1973).
As an example of cube root scaling, let us consider the Oklahoma City Bombing. This event has been reported to have been equivalent to the detonation of
4,000 lbs of TNT at essentially the ground surface. If a location of interest is
100 ft away, the scaled distance is
Z = 100/(4000)1/3 = 6.30 ft/lb1/3
From Figure 6.6 we have an incident pressure of Po = 24.9 psi and a reflected
pressure of Pr = 79.5 psi at this scaled range. This figure also provides the scaled
positive phase duration to /W 1/3 = 1.77 msec/lb1/3 . From this, the total positive
174
BLAST PHENOMENA
Figure 6.6 Blast Parameters for TNT Surface Bursts
phase duration is
to = 1.77 × (4000)1/3 = 28.1 msec
The figure similarly indicates the scaled positive-phase-specific impulse to be
Is = 12.17 psi × msec/lb1/3 and the total impulse is
Is = 12.17 × (4000)1/3 = 193.2 psi × msec
Some structures, such as explosives or propellant processing bays, mail
rooms, and loading docks, often require blast design for containment of internal
PREDICTION OF BLAST PARAMETERS
175
explosion effects. Containment designs must consider that shock waves emanate
from the charge, strike wall and roof components, and reflect to impact other
surfaces. Each surface is subjected to multiple shock waves.
Internal explosions also produce gas and heat, which cause a pressure increase
within the containment. The peak gas pressure, Pg , developed is a function of
the charge weight/free volume ratio in the containment. This pressure buildup is
relatively slow compared to the load duration associated with shock waves. For
simplicity, the peak gas or quasi-static pressure is assumed to rise instantaneously
to the peak gas pressure but is not additive to shock pressures. A curve of peak
gas pressuresversus weight/volume is shown in Figure 6.7.
For example, let us consider the detonation of W = 200 lbs of TNT within
a structure having a free volume of Vf = 50,000 ft3 . This loading density is
W/Vf = 200/50,000 = 4 × 10−3 lb/ft3 . From Figure 6.7 the peak gas pressure is
Pg = 47.2 psi.
The duration of the gas pressures is a function of the vent area available and
the rate at which temperature decays in the containment area. Vent area can
change over time, depending on how fast the vent cover moves away from the
enclosure. The vent velocity is a function of the applied shock impulse and the
frangibility and mass of the vent cover. The duration is determined by first computing the peak gas pressure from Figure 6.7. Then the gas impulse, Ig , is estimated from a set of graphs as a function of the scaled vent area, scaled mass of
the vent cover, loading density, and scaled shock impulse (U.S. Department of
Defense 2008). An equivalent triangular gas duration is then
tg = 2Ig /Pg
Analytical methods have been developed to predict blast loads. These methods fall into two groups: semi-empirical and hydrocode. The semi-empirical approach utilizes a physics-based model to compute selected blast parameters with
coefficients that are “tuned” to match test data. These models are limited to configurations and charge weight ratios for which data are available, but offer the
advantage of quick run times compared with more detailed techniques. Semiempirical codes may offer limited abilities to model shock diffraction, shielding, and reflection. Semi-empirical methods have been developed primarily by
defense-related agencies and are restricted to distribution to the government and
its contractors.
Hydrodynamic codes, or hydrocodes, utilize a grid of computational cells to
track detonation propagation through an explosive charge and shock wave propagation through a medium. Using principles of material behaviors, fundamental thermodynamics, and fluid dynamics, hydrocodes predict pressure, density,
and other key parameters. This type of analysis is much more complex than
the use of empirical relationships. Some of the available tools have user interfaces that greatly simplify the analysis, especially for standard materials, but
considerable effort and expertise are typically required to conduct a competent
analysis.
176
BLAST PHENOMENA
Figure 6.7 Peak Gas Pressure for Internal Detonations
The output from hydrocodes will be in the form of pressure-time series for
selected locations in the model. Application of these loads to the structural
model can be tedious. If the structural analysis involves a time series calculation,
the time varying loads may be transferred into the structural analysis, but the
number of data points to be transferred may be excessive. Normally, a simplification is made to the applied load history to greatly reduce the number of load
points to a series of lines approximating the “real” curve. This reduced multipoint curve is more complex than the linear decay curve used for the empirical
relationship analysis. A coupled analysis may be performed in the hydrocode
PREDICTION OF BLAST PARAMETERS
177
Figure 6.7 (Continued)
which models blast propagation and structural response. This approach offers
additional capability for detailed response analysis but with significant increase
in complexity compared with simulation of blast loads alone.
6.4.2 Bursting Pressure Vessels
Bursting pressure vessel blast predictions are most commonly conducted in one
of two ways: simplified analytical methods or computation fluid dynamics (CFD)
178
BLAST PHENOMENA
models. In either case, it is necessary to select the conditions under which the
vessel would be expected to fail—overpressure/overfilling, runaway reaction, external heating, mechanical or impact load, corrosion, or other mechanisms. This
selection is important because it determines the composition, pressure, temperature, phase, and energy of the vessel contents, all of which can have a significant
effect on the predicted blast loads. In any case, as long as the vessel pressure
is above two atmospheres at failure and the pressure is released abruptly, the
resultant blast will be a shock wave (i.e., suddenly applied).
The most recent simplified method was published by Tang (Tang et al. 1996)
and consists of a series of nondimensionalized curves to predict the side-on pressure and impulse at a particular standoff as a function of pressure ratio (defined
as burst pressure divided by ambient pressure), ratio of specific heats of the vessel contents (a property of the gaseous contents of the vessel), expansion energy
potential of the vessel contents, and the temperature ratio (the ratio of the vessel
internal temperature at failure to the ambient temperature). The vessel energy
term in this analysis should be corrected for ground reflection, liquid fraction in
the vessel, and energy consumed by fragment production. In addition, the sideon pressures need to be modified to account for reflection and clearing effects,
as described in the next chapter.
If more detail is necessary, a CFD model can be applied. These consider the
thermodynamics and gas properties to compute the production and propagation
of blast waves produced by a bursting vessel. It is important to note that there
are complex interactions between the failing vessel and the fluid escaping the
vessel that cannot be adequately modeled by any publicly available current program. In addition, the failure of the vessel is not wholly deterministic, as cracks
can propagate along micro-cracks and grain boundaries in a manner that cannot
usually be predicted. As a result, these numerical analyses will rely heavily on
the experience and expertise of the analyst. The advantage of using a hydrocode
is that it can directly produce load-time distributions on arbitrary surfaces and
account for shielding and focusing effects produced by adjacent structures.
6.4.3 Vapor Cloud Explosions
Prediction of blast loads for vapor cloud explosions can be more complex than
loads for high explosive detonations or bursting pressure vessels. The first step
in determining blast load from vapor cloud explosions is to develop the release
scenario for the flammable material. This selection process is typically done with
input from process engineers and plant operations. Dispersion modeling is performed to model the release scenario and determine the size, shape, and concentration of the flammable vapor cloud. This dispersion is overlaid onto areas of
confinement and congestion to define the flammable mass that may participate
in the explosion. In some cases, cloud dispersion is not evaluated, and the confined/congested volume is conservatively assumed to be filled with a stoichiometric mix of flammable gas. This approach may be justified if it is evident that
credible release scenarios can fill the volume. Blast loads may be determined
PREDICTION OF BLAST PARAMETERS
179
using simplified methods consisting of graphs (blast curves) of pressure and
impulse, or of duration versus scaled standoff, or through more complex modeling using CFD.
In the blast curve method, the scaled standoff is computed by using distance
from the center of the explosion to the point of interest and the energy content of
the confined/congested flammable mass. Scaled pressure and impulse values are
read from blast charts containing flame speed curves. These curves range from
low flame speeds that produce almost negligible blast pressure to the supersonic
flame speed of a detonation producing the highest pressure. Determination of
maximum flame speed requires knowledge of the amount of confinement and
congestion in the flammable portion of the vapor cloud, and the fuel reactivity.
The two most commonly used methods are the Baker-Strehlow-Tang (BST)
(Tang and Baker 1999) and TNO Multi-energy Method (MEM) (Van den Berg
1985). Figure 6.8 shows the overpressure, and Figure 6.9 shows the impulse
as a function of scaled distance and flame speed for BST. To use the curves,
compute the scaled distance Rbar = R(po /E)1/3 in which R = standoff distance,
po = ambient pressure, and E = heat of detonation. Then obtain the scaled pressure Pbar = (P/po ) from Figure 6.8 and unscale to determine the pressure P.
Next, determine the scaled impulse ibar = i ao /(E po )1/3 in which i = impulse and
ao = ambient sound velocity. The impulse, i, follows by unscaling. The equivalent triangular duration is then to = 2i/P.
The process is similar for MEM, using Figures 6.10 and 6.11 for pressure and
duration respectively. Impulse is i = Pt0 /2.
50
Mf=5.2
Mf=4.0
Mf=3.0
Mf=2.0
Mf=1.4
Mf=1.0
Mf=0.7
Mf=0.35
Mf=0.2
10
(P-P0)/P0
1
0.1
0.01
0.002
0.1
1
1/3
(R/E/P0)
Figure 6.8 Peak Pressure for Vapor Cloud Explosions—BST
10
180
BLAST PHENOMENA
Mf=5.2
Mf=4.0
Mf=3.0
Mf=2.0
Mf=1.4
Mf=1.0
Mf=0.7
Mf=0.35
Mf=0.2
i+a0/(E1/3 p02/3)
0.1
0.01
0.001
0.1
1
1/3
(R/E/P0)
10
Figure 6.9 Impulse for Vapor Cloud Explosions—BST
1E+02
10
9
8
7
6
5
4
3
2
1
1E+01
Scaled Pressure (P-P0)/P0
1E+00
1E-01
1E-02
1E-03
1E-04
1E-05
1E-06
1E-07
0.1
1
10
100
1000
Scaled Distance (R(P0/E)1/3)
Figure 6.10 Peak Pressure for Vapor Cloud Explosions—TNO MEM
10000
REFERENCES
181
10
Scaled Duration (t*a*(P0 /E)1/3)
1
2
3
4
1
5
6
7
8
9
10
0.1
0.1
1
10
100
1000
10000
Scaled Distance (R(P0/E)1/3)
Figure 6.11 Duration for Vapor Cloud Explosions—TNO MEM
A different approach is to develop a CFD model of the process area. A grid
of computational cells is constructed, and the controlling combustion chemistry
and gas kinetics equations are solved to track the progression of the reaction and
resulting pressures. The end result is a pressure-time series at selected points.
It is important for the structural analyst to know if the pressure-time plots are
free-field or applied loads. An advantage of this approach is that it allows the
analyst to include the effects of blockage and impingement of objects, which can
substantially influence the dispersion and explosion effects. This capability is not
available in the simplistic methods using scaled energy curves.
6.5 SUMMARY
Blast phenomena are characterized by a rapid release of energy resulting in a
pressure wave that propagates outward from the source. The source can be a high
explosive, propellant, or other gases that are released suddenly. Shock waves
are characterized by an instantaneous rise in pressure from ambient. Pressure
waves, are produced by deflagrations resulting from combustion of low explosives and flammable gases. The characteristics of a blast waves can be predicted
with relatively simple empirical methods or more complex computational fluid
dynamics. Once free-field or side-on blast pressure and impulse are defined, the
applied loads on structures can be determined, as described in the next chapter.
182
BLAST PHENOMENA
REFERENCES
Baker W. E. 1973. Explosions in Air. Austin: University of Texas Press.
Baker, Q.A. and W. E. Baker. 1991. Pros and cons of TNT equivalence for industrial
explosion accidents. Presented at AIChE International Conference on Workshop on
Modeling and Mitigating the Consequences of Accidental Releases of Hazardous Materials, New Orleans, LA, May 21–24, 1991.
Center for Chemical Process Safety. 1994. Guidelines for Evaluating the Characteristics
of Vapor Cloud Explosions, Flash Fires, and BLEVES. New York: American Institute
of Chemical Engineers.
Cooper, W. P. and R. S. Kurowski. 1996. Introduction to the Technology of Explosives.
New York: Wiley-VCH.
Tang M. J. and Q. A. Baker. 1999. A new set of blast curves from vapor cloud explosion.
Process Safety Progress 18 (4): 235–240.
Tang, M. J., C. Y. Cao, and Q. A. Baker. 1996. Calculation of blast effects from bursting vessels.” Presented at the 30th Annual AIChE Loss Prevention Symposium, New
Orleans, LA, February 1996.
U.S. Department of Defense. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense.
Van Den Berg, A. C. 1985. The multi-energy method—A framework for vapor cloud
blast prediction.” Journal of Hazardous Materials, 12 (1): 1–10.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
7
Blast Loading
Paul F. Mlakar and William Bounds
7.1 INTRODUCTION
In the previous chapter, blast phenomena and the characteristics of blast waves
were discussed. In this chapter, methods for determining the dynamic blast loading on structural elements from these characteristics are discussed. Emphasis is
placed on simple approximate procedures that apply widely in practical blastresistant design. More comprehensive methods are also discussed.
7.2 EMPIRICAL METHOD
Although research into the effects of explosions dates back to 1870, most development to determine the blast loading on buildings and other similar structures was started in the 1950s and 1960s by the U.S. military. Several publicly
available military manuals were distributed during this period (TM 5-856 and
TM 5-1300) that presented empirically-derived charts and equations. Several papers and publications published during that period (Newmark 1956, Biggs 1964,
ASCE 1985) also provided information for design.
The empirical method consists of published equations, graphs, tables, and
figures that allow one to determine the principal loading of a blast wave on a
building or a similar structure. Software has also been developed to automate
calculations based on this same source information. The basic advantage of empirical methods is speed and simplicity. Applicable factors and coefficients can
be looked up in tables, blast loading can be calculated manually, and a building
blast load can be determined in a matter of minutes. Empirical methods are also
well suited for situations where the accuracy of the blast source location and
magnitude is uncertain. More comprehensive methods, such as computational
fluid dynamics (CFD), require specialized software, operator training, and, potentially, weeks of data input and verification. The disadvantage of empirical
methods is the lack of flexibility in application. Most data are based on plain
rectangular target structures located in open terrain. Explosions are assumed to
be either an air blast or surface blast. The blast wave is typically based on high
183
184
BLAST LOADING
explosives producing a shock wave, as opposed to a deflagration that produces a
pressure wave. Empirical methods remain popular for the preliminary evaluation
of complex design situations and for the design situations where the location and
magnitude of the blast source are not well understood. The analyst must make a
judgment to pursue an efficient design calculation, and to balance the accuracy
of input information and the desired accuracy of results.
The most extensive and widely referenced publication for empirical design is
UFC 3-340-02 (formerly TM5-1300) (U.S. Department of Defense 2008). This
design manual addresses accidental explosions related to munitions manufacturing, handling, and storage. Nevertheless, many of the procedures are applicable
to buildings and structures designed for other blast scenarios. Basic information from UFC 3-340-02 is provided in this document for external loading. UFC
3-340-02 also provides load determination information on other configurations,
such as partially open cubicles and interior explosions. Information on empirical
methods is available from a number of other sources such as Mays and Smith’s
Blast Effects on Buildings (Mays and Smith 1995), Biggs’s Introduction to Structural Dynamics (Biggs 1964), and ASCE’s Structural Design for Physical Security and Design of Blast Resistant Buildings in Petrochemical Facilities (ASCE
Physical Security 1999, ASCE Petrochemical 1997).
The empirical blast load example calculations provided in this chapter assume
a blast wave interaction with a rectangular structure of finite size, such that the
structure is blast-loaded on all sides. Furthermore, due to the structure depth, or
building dimension parallel to the direction of the blast wave, there will be a significant net lateral force applied to the structure. Mays and Smith describe this
as one of three blast loading situations to keep in mind. The second situation is
a blast wave interacting with a relatively small structure, such as a vehicle, that
is effectively engulfed with blast pressure acting on all sides of the structure at
once. The third situation is for a blast wave acting on a relatively large structure,
such as a large office building, where the magnitude of the blast wave varies significantly across the surfaces of the structure. Some surfaces of these structures
may see little if any external blast loading.
Empirical methods typically rely on a straight-line equivalent of the actual
blast pressure-time relationship, as well as the use of a straight-line equivalent
loading on the structure of interest. The approximation is fairly close, and the
resulting calculations are much simpler.
Details and examples illustrating implementation of empirical methods are included in following sections of this chapter. Key input parameters for the determination of building loads, as presented in Chapter 6, are: the peak side-on overpressure, Pso , the positive phase duration, to , and the shock front velocity, Us .
One other parameter necessary for the determination of building blast loads
is the dynamic wind pressure, qo . Dynamic wind is the movement of air particles resulting from a shock wave. The effect is additive to the blast overpressure
and is a function of the free-field blast overpressure and the obstruction’s shape.
Open-frame structures and small buildings where a blast wave will produce quick
envelopment are most sensitive to dynamic wind. Values of qo can be determined
EMPIRICAL METHOD
185
Figure 7.1 Peak Incident Pressure versus Peak Dynamic Pressure, Density of Air
Behind the Shock Front, and Particle Velocity (UFC 3-340-02)
from Figure 7.1. Alternatively, in the low overpressure range, and at sea level
atmospheric pressure, the following equation from Newmark can be used.
qo = 0.022 (Pso )2
(7.1)
The pressure exerted on a structural element is the dynamic wind pressure
multiplied by a drag coefficient. The drag coefficient, Cd , is a function of the
shape and orientation of the obstructing element. Newmark lists approximate
values of Cd for open-frame structural elements as 2 for structural shapes, 1.25
for box shapes, and 0.8 for cylinders. Values of Cd for enclosed rectangular buildings are provided in the following sections.
186
BLAST LOADING
7.2.1 Empirical Method—Basic Blast Wave Example
The example calculation presented in the following sections is based on the blast
wave and building parameters described below. The given blast wave will be applied normal to the long side of a rectangular flat-roof building. It is further determined that the distance to the explosion and the length of the building are such
that the overpressure and duration do not change significantly over the length of
the building.
Given charge weight, W = 5 tons (TNT) = 10,000 lb:
The charge weight, W , should include any applicable safety factor
Given distance from source to building, R = 215 ft
The blast wave parameters, computed using Figure 6.6 as follows:
To use Figure 6.6, compute the scaled distance:
Z = R/W 1/3 = (215 ft)/(10,000 lb)1/3 = 10.0
Using Z = 10, use Figure 6.6 to determine the following parameters:
Peak side-on overpressure, Pso = 10 psi
Shock front velocity, Us = 1.4 ft/ms
Scaled positive phase duration, to /W 1/3 = 2.6 ms/lb1/3
Compute the positive phase duration:
to = (to /W 1/3 )W 1/3 = (2.6 ms/lb1/3 )(10,000 psi)1/3 = 56 ms
Compute the positive phase wave length:
L w = (Us )to = (1.4 ft/ms)(56 ms) = 78 ft
With Pso = 10 psi, use Figure 7.1 to determine the peak dynamic wind pressure, qo = 2.1 psi
Alternatively, use Equation 7.1 to calculate the peak dynamic wind pressure,
qo = 0.022 (Pso )2 = 0.022 (10 psi)2 = 2.2 psi
Loads are to be computed on an aboveground rectangular building with the
following dimensions,
Building height, B H = 20 ft
Building width, BW = 150 ft
Building depth, B D = 50 ft
7.3 FRONT WALL LOADS
The wall facing the explosion source is subjected to a reflection effect as the
blast wave impacts the facing wall and reflects back towards the blast source.
The reflection effect amplifies the blast pressure on the front or facing side of the
building. Because the overpressure at the top and side edges of the front wall is
FRONT WALL LOADS
vortex
rarefaction
wave
reflected
shock
front
rarefaction
wave
shock front
reflected
shock
front
187
shock front
vortex
Figure 7.2 Blast Wave at Front Wall (TNO Green Book)
less than the reflected overpressure (Figure 7.2), a decay in the reflection effect
takes place that starts at the edges and works inward. The effect is completely
removed after what is called the clearing time, tc . The clearing time is a function
of the height and width of the front wall.
The peak reflected overpressure, Pr , can be determined from Figure 7.3 using
the free-field peak overpressure. Alternatively, for free-field peak overpressures
Figure 7.3 Peak Incident Pressure versus the Ratio of Normal Reflected Pressure/
Incident Pressure for a Free-Air Burst (UFC 3-340-02)
188
BLAST LOADING
less than 40 psi and for sea level atmospheric pressures, the peak reflected overpressure can be determined using the following equation form Newmark:
Pr = [2 + 0.05 (Pso )]Pso
(7.2)
The clearing time can be calculated using the following equation from UFC 3340-02.
tc = 4S/[1 + S/G]Cr
(7.3)
In the preceding equation, S is the lesser of building height or building width.
G is the greater of building height or width, and Cr is the velocity of sound and
can be determined using Figure 7.4.
For the calculation of front wall dynamic wind pressure, a drag coefficient,
Cd , equal to 1.0 is used with the qo value determined from Figure 7.1.
To compute the remainder of the pressure-time curve, the following equations
are used for stagnation pressure, Ps , impulse, Is , and the effective duration, te ,
based on a simplified straight-line approximation. These values are illustrated in
Figure 7.5.
Ps = Pso + Cd (qo )
Is = 0.5 (Pr − Ps )tc + 0.5Ps to
te = 2Is /Pr
(7.4)
(7.5)
(7.6)
The blast wave angle of incidence affects the blast pressure load on the front
wall. This angle is taken as 0◦ for a blast wave traveling perpendicular into the
plane of the front wall where the full reflected overpressure is applied, and taken
as 90◦ for a blast wave traveling parallel to a surface, where the free-field, or sideon overpressure is applied. For intermediate values of the angle of incidence,
Figure 7.6 provides coefficients to calculate the applied pressure for use in the
following equation.
Pr α = (Cr α )(Pso )
(7.7)
7.3.1 Empirical Method—Front Wall Loading Example
The front wall is assumed to span vertically from foundation to roof. The calculation will be for a typical wall segment one foot wide.
With Pso = 10 psi, use Figure 7.3 to determine the reflected overpressure
ratio, Pr /Pso = 2.5.
Pr = (Pr /Pso )Pso = (2.5)(10 psi) = 25 psi
189
Figure 7.4 Velocity of Sound in Reflected Overpressure Region versus Peak Incident Overpressure (UFC 3-340-02)
190
BLAST LOADING
P
Pr
Equivalent Loading
Ps
tc
Figure 7.5
te
td
t
Front Wall Loading Diagram
Alternatively, use Equation 7.2 to compute the reflected overpressure,
Pr = [2 + 0.05 (Pso )]Pso = [2 + 0.05(10 psi)](10 psi) = 25 psi
To compute the reflected pressure clearing time, two parameters are needed,
S = minimum of B H or BW = minimum of 20 ft or 150 ft = 20 ft
G = larger of B H or BW = larger of 20 ft or 150 ft = 150 ft
With Pso = 10 psi, use Figure 7.4 to determine the sound velocity in the reflected region, Cr = 1.28 ft/ms
Use Equation 7.3 to compute the reflected overpressure clearing time,
tc = 4S/[1 + S/G]Cr = 4(20 ft)/[1 + (20 ft)/(150 ft)](1.28 ft/ms) = 55 ms
This result is nearly equal to to , thus tc ≈ to = 56 ms
Drag coefficient, Cd = 1.0
Use Equation 7.4 to calculate the stagnation pressure,
Ps = Pso + Cd (qo ) = (10 psi) + (1.0)(2.2 psi) = 12.2 psi
Use Equation 7.5 to calculate the front wall impulse,
Is = 0.5 (Pr − Ps )tc + 0.5Ps to = 0.5 [(25 psi) − (12.2 psi)](56 ms)
+ 0.5 (12.2 psi)(56 ms) = 700 psi-ms
Use Equation 7.6 to calculate the effective duration,
te = 2Is /Pr = 2 (700 psi-ms)/(25 psi) = 56 ms
pressure
25 psi
56 ms
time
191
Cr␣ = Pr␣/Pso
0
2
4
6
8
10
12
14
0
10
30
40
50
Angle of incidence, ␣ (Degrees)
60
70
Figure 7.6 Reflected Pressure Coefficient versus Angle of Incidence (UFC 3-340-02)
20
Peak Incident Overpressure, psi
5000
70
3000
50
2000
30
1000
20
500
10
400
5
300
2
200
1
150
0.5
100
0.2
80
90
192
BLAST LOADING
7.3.2 Empirical Method—Oblique Angle Example
Given angle of incidence, α = 40 degrees
Given side-on overpressure, Pso = 10 psi
Using Figure 7.6, determine the reflected pressure coefficient, Cr α = 2.4
Using Equation 7.7, calculate the reflected overpressure,
Pr α = (C1α )(Pso ) = (2.4)(10 psi) = 24 psi
7.4 SIDE WALL AND ROOF LOADS
Roofs and side walls represent surfaces that are parallel to the path of the advancing blast wave. There is no reflection effect for this situation; however, the
average pressure applicable to a specific area, for example a structural element,
depends on the length of the blast wave and the length of the structural element.
Figure 7.7 shows a diagram of a blast wave traveling across a roof. If a section of
the roof tributary to a supporting beam is oriented perpendicular to the traveling
wave, then the roof length would be short and the blast wave would have a full
effect. If the roof beam were oriented parallel to the traveling wave, then the roof
length would be longer, the average pressure would be reduced, and a rise time
would become evident. Calculations for these two situations are presented in the
side wall and roof examples. Values are obtained as follows:
r Equivalent uniform pressure factor, CE , is determined from Figure 7.8.
r Scaled uniform pressure rise time, tr , is determined from Figure 7.9.
r Scaled uniform pressure duration, to , is determined from Figure 7.10.
A building’s depth could be significant with respect to the distance from the
blast source. Thus the peak free-field overpressure for a side or rear surface could
be lower than that calculated for the front wall. The calculation of peak free-field
SHOCK FRONT
AT TIME t
f
b
Lw
L
Figure 7.7 Roof Loading Diagram
SIDE WALL AND ROOF LOADS
193
1
0.9
Positive Pressure, CE = PR/Psof
Negative Pressure, CE = PR/Psof
0.8
0.7
0.6
Equivalent Load Factor
0.5
0.4
0.3
Positive
0.2
Negative
0.1
0.2
0.3
0.4 0.5 0.60.70.8 1
2
3
Wave length/Span length, Lwf/L
4
5
6 7 8 9 10
Figure 7.8 Peak Equivalent Uniform Roof Pressure (UFC 3-340-02)
overpressure is determined using the criteria in Chapter 6. In many cases this
effect is negligible and is ignored, for simplicity.
For the calculation of roof and side wall dynamic wind pressure, a drag coefficient, Cd , determined from Table 7.1 is used with the qo value determined from
Figure 7.1.
194
BLAST LOADING
5
Numbers next to curves indicate
peak incident overpressure. psi
4
3
2
Scaled Rise Time
2
4
8
1
0.9
0.8
0.7
16
0.6
0.5
32
0.4
0.3
0.2
0.5
0.6
0.7
0.8
0.9
1
2
Wave length/Span length, Lwf/L
Figure 7.9 Scaled Rise Time of Equivalent Uniform Positive Roof Pressures (UFC 3340-02)
The total effective pressure on a roof or side wall is determined from the
following,
(7.8)
Pa = C E Pso + Cd qo
7.4.1 Empirical Method—Side Wall Loading Example
The side wall is the same as the front wall, spanning vertically from foundation
to roof. This calculation will be for a unit width wall segment, L = 1 ft.
SIDE WALL AND ROOF LOADS
195
Figure 7.10 Scaled Duration of Equivalent Uniform Roof Pressures (UFC 3-340-02)
196
BLAST LOADING
Table 7.1 Roof, Side Wall, and Rear Wall
Drag Coefficients
Peak dynamic pressure
Drag coefficient
0–25 psi
25–50 psi
50–130 psi
−0.40
−0.30
−0.20
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is needed,
L w /L = (78 ft)/(1 ft) = 78
With such a short wall segment in comparison to the blast wave length, it isn’t
necessary to obtain values from Figures 7.8, 7.9, and 7.10.
By inspection, C E = 1.0, use Equation 7.8 to calculate the effective peak
overpressure,
Pa = C E Pso + Cd qo = (1.0)(10 psi) + (−0.4)(2.2 psi) = 9.1 psi
By inspection, the effective rise time, tr = 0.0 ms
By inspection, the effective duration, td = 56 ms
pressure
9.1 psi
56 ms
time
7.4.2 Empirical Method—Roof Loading Example
The roof is a slab spanning between roof beams, and an appropriate span should
be used for L. However, to illustrate the use of the figures, a section spanning the
entire building will be used, thus L = 50 ft
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is needed,
L w /L = (78 ft)/(50 ft) = 1.6
With L w /L = 1.6, use Figure 7.8 to determine the effective uniform pressure
factor, C E ≈ 0.60
REAR WALL LOADS
197
Use Equation 7.8 to calculate the effective peak overpressure,
Pa = C E Pso + Cd qo = (0.60)(10 psi) + (−0.4)(2.2 psi) = 5.1 psi
With L w /L = 1.6, use Figure 7.9 to determine the effective scaled rise time,
tr /W 1/3 ≈ 1.4 ms/lb1/3
Rise time, tr = (tr /W 1/3 )(W 1/3 ) = (1.4 ms/lb1/3 )(10,000 lb)1/3 = 30 ms
With L w /L = 1.6, use Figure 7.10 to determine the effective scaled duration,
td /W 1/3 ≈ 3.0
Duration, td = (td /W 1/3 (W 1/3 ) = (3.0 ms/lb1/3 )(10,000 lb)1/3 = 65 ms
Total positive phase duration
to = tr + td = (30 ms) + (65 ms) = 95 ms
7.5 REAR WALL LOADS
The rear wall is the wall facing directly away from the blast source, as illustrated
in Figure 7.11. The calculation of rear wall blast loads is similar to that for the
side walls or roof. In fact, UFC 3-340-02 treats the rear wall as a roof extension.
For the calculation of rear wall dynamic wind pressure, a drag coefficient, Cd ,
determined from Table 7.1, is used with the qo value determined from Figure 7.1.
7.5.1 Empirical Method—Rear Wall Loading Example
The example will be for a wall segment 1 foot wide, L = building height = 20 ft
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is
needed,
L w /L = (78 ft)/(20 ft) = 3.9
shock front
shock front
vortex
diffracted
shock
front
diffracted
shock
front
vortex
Figure 7.11 Rear Wall (TNO Green Book)
198
BLAST LOADING
With L w /L = 3.9, use Figure 7.8 to determine the effective uniform pressure
factor, C E ≈ 0.85
Use Equation 7.8 to calculate the effective peak overpressure,
Pa = C E Pso + Cd qo = (0.85)(10 psi) + (−0.4)(2.2 psi) = 7.6 psi
With L w /L = 3.9, extrapolate from Figure 7.9 to approximate the effective
scaled rise time, tr /W 1/3 ≈ 0.5 ms/lb1/3
rise time, tr = (tr /W 1/3 )(W 1/3 ) = (0.5 ms/lb1/3 )(10,000 lb)1/3 = 11 ms
Note that an alternative rise time calculation, based on the time for the shock
front to travel across the element, would be tr = L/Us = (20 ft)/(1.4 ft/ms) =
14 ms
With L w /L = 3.9, extrapolate from Figure 7.10 to approximate the effective
scaled duration, td /W 1/3 ≈ 2.1
Duration, td = (td /W 1/3 (W 1/3 ) = (2.1 ms/lb1/3 )(10,000 lb)1/3 = 45 ms
Note that the duration approaches the free field duration as L w /L gets large.
Total positive phase duration,
to = tr + td = (11 ms) + (45 ms) = 56 ms
pressure
7.6 psi
11 ms
56 ms
time
7.6 CONFINED EXPLOSIONS
As discussed in Chapter 6 an interior explosion involves more phenomena than
an external explosion, and the loading (Figure 7.12) changes accordingly. In the
internal explosion, there is first the directly incident shock as in an external explosion. This is followed by multiple shock reflections off the other surfaces
that are confining the explosion. Finally, there is a longer duration gas pressure
throughout the interior as the gaseous products of the detonation come to thermodynamic equilibrium.
In practice, it is often convenient to approximate this typical loading with a
suddenly applied bilinear pulse, as shown in Figure 7.13. The first part of this
loading is the directly incident shock as computed in Section 7.3 for front wall
loads. The second part is a triangular approximation to the gas pressure, which
generally has a lower peak but longer duration than the shock loading.
CONFINED EXPLOSIONS
199
Figure 7.12 Loading from Confined Explosion
To calculate this gas pressure loading, the peak gas pressure, Pg , is obtained from Figure 6.7, as described in the previous chapter. Then the scaled
gas impulse, i g /W 1/3 , is computed from Figures 7.14 through 7.25 as a function of loading density, W/V f , the scaled reflected shock impulse, ir /W 1/3 , and
the scaled mass, W F /W 1/3 , of the frangible openings in the confined volume.
Table 7.2 indicates which of the Figures 7.14 through 7.25 pertains to various
IDEALIZED SHOCK
PRESSURES
PRESSURE
ir = REFLECTED IMPULSE
IDEALIZED GAS
PRESSURES
TIME
Figure 7.13 Approximate Loading from Confined Explosion
200
BLAST LOADING
Table 7.2 Figures for Scaled Gas Impulse ig/W 1/3
W/Vf
W/Vf
W/Vf
W/Vf
= 0.002
= 0.015
= 0.15
= 1.0
ir /W 1/3 = 20
ir /W 1/3 = 100
ir /W 1/3 = 600
ir /W 1/3 = 2,000
Fig. 7.14
Fig. 7.17
Fig. 7.20
Fig. 7.15
Fig. 7.18
Fig. 7.21
Fig. 7.23
Fig. 7.16
Fig. 7.19
Fig. 7.22
Fig. 7.24
Fig. 7.25
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
2/3
A/Vf
Figure 7.14 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 20
10
201
CONFINED EXPLOSIONS
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
10
2/3
A/Vf
Figure 7.15 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 100
values of W/V f and ir /W 1/3 . For values between those given in this table, the
scaled gas impulse may be found by interpolation. Once the scaled gas impulse
is obtained, the actual impulse follows by unscaling, i.e., multiplication by W 1/3 .
The equivalent triangular duration of the gas loading is then tg = 2ir /Pg .
For example, let us estimate the loading from the detonation of W = 16 lbs
of TNT at the center of the floor a cubicle that measures 20 ft on a side. Suppose
202
BLAST LOADING
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
100
3
1
10
0.3
0
1
0.01
0.1
SCALED VENT AREA,
1
10
A/Vf2/3
Figure 7.16 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 600
that the total vent area is A = 20 ft2 and the weight of the frangible surfaces is
W F = 25 lbs.
Using the method described in Section 7.3, the directly incident shock is described by
Pr = 317 psi
to = 0.868 msec
203
CONFINED EXPLOSIONS
10000
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
1
0.1
SCALED VENT AREA,
A/Vf2/3
Figure 7.17 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 20
For the gas pressure, the loading density is
W/V f = 16/203 = 0.002 lbs/ft3
From Figure 6.7, the corresponding peak pressure is
Pg = 25.5 psi
10
204
BLAST LOADING
10000
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
1
100
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
10
A/Vf2/3
Figure 7.18 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 100
For the scaled reflected shock impulse, ir /W 1/3 = 54.6 psi × msec/lb1/3
Interpolation between Figures 7.14 and 7.15 gives the scaled gas impulse
ig /W 1/3 = 1990 psi × msec/lb1/3
The actual gas impulse is then
i g = 1990 × 161/3 = 5010 psi × msec
205
CONFINED EXPLOSIONS
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
10000
wr /w1/3
1000
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.19 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 600
and the equivalent triangular duration is
tg = 2 × 5010/25.5 = 393 msec
This dynamic loading is summarized in Figure 7.26. As to tg in this case,
the shock portion appears very close to the vertical axis.
206
BLAST LOADING
10000
wr /w1/3
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.20 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 20
7.7 LEAKAGE
If the openings in the exterior of a structure are not designed for an expected
blast loading, there will be a leakage of pressure into the interior. Idealized
pressure-time series for this loading are discussed in UFC 3-340-02 as a function
207
LEAKAGE
10000
wr /w1/3
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.21 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 100
of the structural geometry and shock wave characteristics. Different curves are
developed for the exterior front wall, interior front wall, interior side wall and
roof, and interior back wall. This manual also discusses closures that are deliberately designed to resist the blast but have small openings for vents and ducts
that admit some overpressure.
208
BLAST LOADING
10000
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
100
1
0.3
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.22 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 600
7.8 RAY-TRACING PROCEDURES
Ray-tracing algorithms have been developed to provide a higher resolution of
the blast loading than provided by the previously discussed empirical methods in complex geometrical situations. These procedures generally work backwards from each location of interest on a structure to capture a finite number
209
RAY-TRACING PROCEDURES
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
30
1000
10
3
1
0.3
100
10
0.01
0
0.1
SCALED VENT AREA,
1
10
2/3
A/Vf
Figure 7.23 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 100
of the possible multiple reflections at each location. As more reflections are
considered, the fidelity increases, but so does the computational effort. In practice, the coefficients of such physics-based models are “tuned” to match test
data. Accordingly, these are often termed “semi-empirical” models. Their applicability is limited to configurations and charge weight ratios for which data are
available. Some semi-empirical codes have the capability to approximate shock
210
BLAST LOADING
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
30
1000
10
3
1
100
0.3
10
0.01
0
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.24 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 600
diffraction, shielding, and reflection. In both fidelity and computational effort,
these methods lie between the previously discussed fully empirical procedures
and fully physics-based computational fluid dynamics. These semi-empirical
methods have been developed primarily by defense-related agencies and are restricted in distribution to the government and its contractors.
211
RAY-TRACING PROCEDURES
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
1000
30
10
3
100
1
0.3
10
0.01
0
0.1
1
2/3
SCALED VENT AREA, A/Vf
Figure 7.25 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 2,000
10
212
BLAST LOADING
Figure 7.26 Approximate Loading for Detonation of 16 lbs of TNT in a 20-ft Cubicle
7.9 SUMMARY
In this chapter, methods for determining the dynamic blast loading on structural
elements from the characteristics of blast phenomena have been discussed and
illustrated with examples. Emphasis was placed on simple approximate procedures that apply widely in practical blast-resistant design. More comprehensive
methods were also discussed.
REFERENCES
ASCE. 1985. Design of Structures to Resist Nuclear Weapons Effects (Manual No. 42).
Reston, VA: American Society of Civil Engineers.
ASCE Petrochemical. 1997. Design of Blast Resistant Buildings in Petrochemical Facilities. Reston, VA: Task Committee on Blast Resistant Design, American Society of
Civil Engineers.
ASCE Physical Security. 1999. Structural Design for Physical Security, Reston, VA:
American Society of Civil Engineers.
Biggs J. M. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill Book
Company.
Committee for the Prevention of Disasters Due to Dangerous Substances. 1992. Method
for the determination of possible damage to people and objects resulting from releases
of hazardous materials (CPR 16E) TNO Green Book. The Hague, The Netherlands:
The Director-General of Labour.
REFERENCES
213
Mays, G. C. and P. D. Smith, eds. 1995. Blast Effects on Buildings. London: Thomas
Tedford Publications.
Newmark, Nathan M. 1956. An engineering approach to blast-resistant design (Paper
2786). ASCE Transactions 121: 45–64. Reston, VA: American Society of Civil Engineers.
U.S. Department of Defense. 1965. Design of Structures to Resist the Effects of Atomic
Weapons (TM 5-856). Washington, DC: U.S. Department of Defense.
. 1990. Structures to Resist the Effects of Accidental Explosions (TM 5-1300).
Washington, DC: U.S. Department of Defense.
. 2008. Structures to Resist the Effects of Accidental Explosions
(UFC 3-340-02). Washington, DC: U.S. Department of Defense.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
8
Fragmentation
Kim King
8.1 INTRODUCTION
Injury and structural damage during an explosion may result not only from the
direct pressure and impulse of an air blast, but also from the impact of flying
objects called debris and fragmentation. Military weapons are typically explosive
charges with some type of metal casing. Upon detonation, this case is ruptured
and expelled as fragmentation at high velocity. Similarly, terrorist devices may be
embedded with objects such as ball bearings or nails that will be ejected at high
velocity. Other fragmentation may occur as objects interact with the blast wave
created from a detonation and become airborne. Fragments can range in size
from small irregularly shaped objects (fractions of an inch) to large objects such
as portions of buildings or vehicles. Two types of fragments may be generated
during an explosion: primary and secondary fragments.
Fragmentation generated during an explosion can damage structures by penetrating walls as well as impacting structural elements with sufficient impulse to
cause material failure or global structural failure of a structural element. This
chapter discusses fragmentation resulting from an explosion and established
methods of predicting the hazards of airborne fragments.
8.2 DEBRIS
Debris generically refers to the broken remains of a destroyed object. Debris
from a destroyed structure near an explosion can be thrown as secondary fragmentation.
8.3 LOADINGS
Fragmentation can typically be separated into two categories: primary and secondary fragmentation. Primary fragments are typically produced from the container or casing of an explosive charge. Primary fragments can also be generated in manufacturing processes where equipment or tools are in contact with
explosive material. Examples of primary fragments include the casing of
215
216
FRAGMENTATION
conventional munitions, metal containers such as kettles and hoppers used in
the manufacture of explosives, the steel casing of pipe bombs, and the metal
housing of rocket engines. For high explosives, the casing will typically shatter
into many small fragments traveling at very high initial velocities up to several
thousand feet per second. Primary fragments are typically small, on the order of
1 gram. They typically have “chunky” or irregular geometry with linear dimensions all of the same order. However, some casings are scored or formed to fail
in a specific manner to generate a specific fragment pattern.
Secondary fragments are produced when the shock wave encounters objects
or structures located near the source of detonation. Secondary fragments are also
known as “secondary debris” or “secondary missiles.” Secondary fragments are
generally larger than primary fragments and have lower initial velocities. Examples of secondary fragments include broken glass from window systems, debris from exterior walls, and dislodged architectural ornamentation. Secondary
fragmentation can also include items located near (but not in contact with) the
explosive source, such as parts of an automobile near a terrorist vehicle bomb.
Damage from secondary fragments is not easy to accurately predict because of
the variation in size, shape, and initial velocity.
8.3.1 Primary Fragmentation
Much work has been done to characterize the type of fragments and corresponding fragment velocity of military munitions. Many factors can affect the primary
fragment shape, mass, and velocity, including the explosive type, shape of the explosive charge, arrangement of casing, and so forth. For simplification, the most
common shape of a cased explosive (a cylinder) with uniform case thickness will
be considered.
Primary Fragment Shape The shape of primary fragments may be predictable
if, for example, the casing is scored to produce a particular shape, or a specific
munition, with a known fragment shape, is known. But a standard shape for
primary fragments is typically assumed for design purposes where a specific
shape is not known. The typical fragment shape is shown in Figure 8.1.
Primary Fragment Mass For the most common device shape, a uniform cylinder, and a given confidence level (CL ≤ 0.999), the design fragment weight can
be estimated with the following procedure:
Wd f = M A2 · ln2 (1 − C L )
where: Wdf = design fragment weight (ounces)
MA = fragment distribution factor
CL = confidence level
(8.1)
217
LOADINGS
d = Diameter of cylindrical portion of
fragment
r
r
d
r = Radius of hemispherical portion
of fragment (d/2)
r
r
r
d
Wf = Fragment weight, 0.645.d 3/γ = 0.186.d 3
D = Caliber density = 0.186.lb/in3
N = Nose shape factor = 0.72 + 0.25.(n – 0.25)1/2 = 0.845
n = Caliber radius of the tangent ogive of the fragment nose = r/d = 0.5
Figure 8.1 Standard Primary Fragment Shape
1/3
M A = B · tc5/6 · di
· (1 + tc di )
(8.2)
where: B = explosive constant (see Table 8.1)
tc = average casing thickness (inch)
di = average inside diameter of casing (inch)
Table 8.1 Mott Scaling Constants for Mild Steel Casings and Various
Explosives (Mott 1947)
A (oz1/2 inches−3/2 )
B (oz1/2 inches−7/6 )
Baratol
—
0.512
Composition A-3
—
0.22
Composition B
0.214
0.222
0.197
Explosive
Cyclotol (75/25)
—
H-6
—
0.276
HBX-1
—
0.256
HBX-3
—
0.323
Pentolite (50/50)
0.238
0.248
PTX-1
—
0.222
PTX-2
—
0.227
RDX
0.205
0.212
Tetryl
0.265
0.272
TNT
0.302
0.312
218
FRAGMENTATION
Table 8.2 Gurney Energy
Specific Weight (lb/inch3 )
Explosive
√
2 · E
Composition B
0.0621
9100
Composition C-3
0.0578
8800
HMX
0.0682
9750
Nitromethane
0.0411
7900
PBX-9404
0.0664
9500
PETN
0.0635
9600
RDX
0.0639
9600
TACOT
0.0581
7000
Tetryl
0.0585
8200
TNT
0.0588
8000
Trimonite No. 1
0.0397
3400
Tritonal (TNT/AL - 80/20)
0.0621
7600
Primary Fragmentation Velocity The most common technique for calculating
the initial velocity of primary fragmentation in contact with an explosive charge,
the Gurney method, is provided here (Gurney 1947). The initial velocity of primary fragments produced from a cylindrical cased explosive is a function of the
explosive output and the ratio of the explosive charge weight to casing weight.
The initial velocity of primary fragments resulting from a cylindrical cased explosive with evenly distributed explosives and casing is expressed as:
Vop
√
= FS · 2 · E ·
W
Wc
1 + 0.5 · W Wc
(8.3)
where: √ Vop = initial velocity of primary fragment (feet/sec)
2 · E = Gurney Energy Constant (see Table 8.2)
FS = factor of safety (1.2)
W = design charge weight = FS·Wact (feet/sec)
Wact = actual quantity of explosive material (lb)
Wc = weight of casing (lb)
8.3.2 Secondary Fragmentation
Secondary fragments generated during an explosion can represent a wide variety of sizes, shapes, and initial velocities. An array of scenarios must typically
be considered to establish the design fragment (worst-case loading on the structure). Varying factors such as fragment size, shape, structural constraint, and
orientation must be evaluated to establish the controlling fragment scenario that
produces the most severe response of the structure in question.
LOADINGS
219
Pressure
Pr
CD.Q
F (t) = CD.q(t) = CD.1/2.ρu 2
ta
t1
t2
t3
Time
Figure 8.2 Idealized Time Loading on an Irregular Fragment
Initial steps for identifying the hazard associated with secondary fragmentation are listed below.
1. Define blast loading (see Chapter 7 for blast loading).
2. Determine distance from the center of the explosive to the potential fragment (refer to structural details of the equipment and/or structure in consideration).
3. Determine the size, shape, and connection of the potential fragment (see
structural details for the equipment and/or structure in consideration).
4. Determine potential fragment velocity using the methods detailed in this
chapter.
The blast load applied to the object can be simplified to a three-step function
that represents the rise to peak pressure, initial pressure decay that is affected by
diffraction of the wave around the object and rarefaction waves that form across
the front of the object, and finally the drag phase loading (see Figure 8.2).
Unconstrained Fragments Unconstrained fragments are generated when objects that are not attached interact with the blast wave and are thrown from the
incident origin. Predictions for unconstrained fragment velocity assume that the
object is not attached (i.e., no energy lost breaking the object free), the object
will not deform, and gravity will not affect velocity during acceleration.
220
FRAGMENTATION
Unconstrained Fragments Far from the Charge Potential fragments are defined
as far from the charge if the objects are located at a distance more than 20 times
the radius of the explosive. For potential fragments located far from the charge,
the initial fragment velocity can be calculated using the following methods (U.S.
Department of Defense 2008).
The basic equation of motion describes the object acceleration, where the
time-dependent pressure applied to the area of the object facing the blast is equal
to the mass of the object times acceleration. This relationship can be integrated
and rearranged to yield:
td
V (td ) =
A
id
M
(8.4)
0
where: V = velocity (feet/sec)
td = load duration (msec)
A = area of object facing blast (inches2 )
M = mass of object (lb-msec2 /inch)
id = total drag and diffraction (psi-msec/lb1/3 )
Equation 8.4 can be explicitly resolved if the loading history can be mathematically defined in a function. However, Figure 8.3 can be used to graphically
solve for the nondimensional velocity and then solve for the object initial velocity, by solving for the quantity Pso /Po (on the ordinate) and the following value
on the abscissa:
12 · C D · i s · ao
1000 · Pso · (K · H + X )
(8.5)
Establish the nondimensional object velocity using the various curves. The
nondimensional velocity (¯v ) is equal to:
v¯ =
144 · vo · M · ao
106 · Po · (K · H + X )
(8.6)
Rearranging Equation 8.6 will result in the initial object velocity.
where: Pso = peak incident overpressure (psi)
Po = atmospheric pressure (psi)
Cd = drag coefficient (see Figure 8.4)
is = incident-specific impulse (psi-msec)
ao = velocity of sound in air (feet/sec)
K = constant (see Table 8.3)
H = minimum transverse dimension of the mean presented area of
object (inch)
LOADINGS
221
5
1
PsoPo
0.1
50.0
0.01
10.0
5.0
1.0
0.5
0.001
0.0001
2E-5
0.2
0.1
0.05
Numbers adjacent to curves indicate
nondimensional object velocity =
(144 vo.M.ao)/106.Po.A.(K.H+X)
1
10
0.01
0.005
0.001
100
1,000
10,000
100,000
1,000,000
(12.Cd.js.ao)/(1000.Pso.(K.H +X))
Figure 8.3 Nondimensional Object Velocity
X = distance from the front of the object to the location of its largest
cross section normal to the plane of the shock front (inch)
vo = initial velocity of object (feet/sec)
Unconstrained Fragments Near the Charge Potential fragments are defined as
near the charge if the objects are located at a distance less than or equal to 20
times the radius of explosive. Initial fragment velocity for potential fragments
can be calculated using the following methods (U.S. Department of Defense
2008).
The first step to determine the velocity of a close-in unconstrained fragment
is to determine the specific impulse. Two charge configurations are considered,
spherical and cylindrical. For spherical charges with R/Re ≤ 5.07:
i
=
β · Reff
Re
Rt
0.158
Re
= 38,000 ·
R
1.4
(8.7)
and Reff = Re
For cylindrical charges with R/Re ≤ 5.25:
i
=
β · Reff
Re
Rt
0.158
Re
= 46,500 ·
R
(8.8)
222
FRAGMENTATION
Shape
Sketch
CO
FLOW
Circular cylinder (Long rod),
side-on
1.20
FLOW
0.47
Sphere
FLOW
0.82
Rod, end-on
FLOW
OR
Disc, face-on
FLOW
Cube, face-on
Cube, edge-on
FLOW
Long rectangular member,
face-on
FLOW
Long rectangular member,
edge-on
FLOW
1.17
1.05
0.80
2.05
1.55
FLOW
Narrow strip, face-on
Figure 8.4 Drag Coefficients (Baker et al. 1977)
1.98
LOADINGS
223
Table 8.3 K Constant
Steel
K-Value
Object on reflecting surface (i.e., on ground)
4
Object in air
2
For cylindrical charges with 5.25 < R/Re ≤ 10:
i
=
β · Reff
Re
Rt
0.158
Re
= 161,700 ·
R
1.75
(8.9)
and
Reff
where:
Le
= 0.909 ·
Re
0.333
· Re
(8.10)
i = specific acquired impulse (psi-msec)
β = nondimensional shape factor of the target from Figure 8.5
Reff = effective radius (feet)
Re = radius of the explosive (feet)
R = standoff distance (feet)
Rt = target radius (feet)
Le = length of cylindrical explosive (inch)
Figure 8.6 graphically shows specific acquired impulse for spheres and cylinders. The data for this relationship were experimentally derived and should not
be used beyond the distances shown in the figure. When these standoff distances
are greater than plotted values, the normal reflected impulse may be used as an
approximation for specific acquired impulse.
Because the target is so close to the charge, the actual load history is unimportant, and the impulse acting on the target is equal to the applied momentum.
Therefore, i=M·V/A and the velocity of a specific target is:
vo =
1000 · i · β · A
12 · M
(8.11)
Constrained Fragments A constrained secondary fragment is generated when
an object is torn loose from its moorings and thrown from the explosion source.
The velocity of a constrained secondary fragment is dependent on the specific
impulse applied to the object minus the impulse required to free the object from
the support. Therefore, the following is true:
I − Ist = M · vo
(8.12)
224
FRAGMENTATION
Cylindrical Explosive
2Re
2Rt
Le
a)
b)
Exposed Flat
Face β = 1.0
Exposed Cylindrical
Surface β = π/4
c)
Exposed Spherical
Surface β = 2/3
Spherical Explosive
2Re
2Rt
R
Figure 8.5 Nondimensional Shape Factor (U.S. Department of Defense 2008)
50,000
Specific Impulse
Cylindrical Charge
Spherical Charge
10,000
5,000
2,000
1
5
R/Re
Figure 8.6 Specific Acquired Impulse
10
LOADINGS
Table 8.4
225
Material Toughness
Toughness inches-lbf/inches3
Steel
ASTM A36
12,000
ASTM A441
15,000
ASTM A514 Grade F
19,000
where: I = total blast impulse applied to the fragment (psi-msec)
Ist = impulse required to free the fragment from the support (psi-msec)
Empirical methods have been produced to estimate the initial fragment velocity from a cantilever or a fixed-fixed beam (clamped-clamped) (U.S. Department
of Defense 2008). If the following relationship is true, the object velocity is zero
because the applied impulse is insufficient to disengage the object.
If:
i · bf
≤ 0.602,
2 · L b 0.3
ρ f · T · Ab ·
bf
where:
(8.13)
i = unit impulse applied to the object (psi-msec)
bf = width of fragment exposed to blast (inch)
ρ f = mass density of the fragment lb-msec2 /inch4 )
Ab = smallest cross-sectional area of the fragment (inch2 )
Lb = length of fragment exposed to blast (inch)
T = toughness of material (see Table 8.4)
Then, vo = 0 · ft s [m/s]; otherwise:
⎡
⎛
⎢
⎜
1000
⎢
⎜
T
·
·
+
K
·
vo =
K
⎢ 1
2 ⎜
ρf
⎣
⎝
12
⎞⎤
⎟⎥
i · bf
⎟⎥
0.3 ⎟⎥ (8.14)
⎠⎦
2 · Lb
ρ f · T · Ab ·
bf
where: K 1 = constant (see Table 8.5)
K 2 = constant (see Table 8.5)
Table 8.5
Constants K1 and K2
Support Condition
K1
K2
ASTM A36
−0.2369
+0.3931
ASTM A441
−0.6498
+0.4358
226
FRAGMENTATION
4
Cantilever
Clamped-Clamped
3
.
v2 = 12 V . ρf .T
1000 √
2
1
0
0
1.5
3
4.5
2.Lb
i .bf .
6
7.5
9
10.5
0.3
bf
Ab ρt .T
√
Figure 8.7 Scaled Fragment Velocities for Constrained Objects
Figure 8.7 shows Equation 8.14 graphically for both cantilever and fixed-fixed
support conditions.
8.4 DESIGN FRAGMENT PARAMETERS
As a fragment is thrown from its initial position, several factors influence the
impact the fragment will have on the target. Those factors include the fragment
trajectory, the velocity upon impact, and the angle at which the fragment actually
impacts the target (referred to as angle of obliquity). A zero angle of obliquity is
the most conservative and typically assumed for design, since the actual angle of
obliquity is rarely known.
8.4.1 Fragment Final Velocity
Fragment velocities vary with distance due primarily to drag forces. As the fragment travels away from the event, the initial velocity will decay until impact.
It is assumed the initial velocity is achieved instantaneously. If a target is located close to an explosive event (less than 20 feet [U.S. Department of Defense
2008]), the decay in velocity is minimal and can be ignored. For targets farther
from the event, the following method can estimate the final fragment velocity.
Vs = vo · e−[12·kv ·R f ]
(8.15)
DESIGN FRAGMENT PARAMETERS
227
where: Vs = fragment velocity at distance of interest (Rf )
Rf = distance from the center of detonation
kv = velocity decay coefficient
and:
kv =
where:
A
· ρa · C D
Wf
(8.16)
Wf = fragment weight (ounces)
A/Wf = fragment form factor
ρ a = specific density of air (ounces/inch3 )
For primary fragments, some basic assumptions simplify the equation for final
fragment velocity.
Assume:
1/3
A/W f = 0.78/W f (for mild steel fragments)
ρa = 0.00071−ounces/inch3
C D = 0.6
Resulting in:
Vs = vo · e
− 0.004·
Rf
W f 1/3
(8.17)
8.4.2 Fragment Trajectory
Once the initial velocity of primary or secondary fragments is determined, it
will be necessary to establish the potential range of the accelerated fragments, to
establish the hazard associated with the fragments. The range is established by
solving the equations that define the trajectory the fragment will follow until it
strikes the ground or a target. Because fragments generated during an explosion
are typically irregular in shape and mass, determining the fluid dynamic forces
acting on the fragments to influence the trajectory can be difficult, and simplified
fluid dynamic force prediction methods are typically adopted. Baker describes
the acceleration in the X and Y directions based on both lift and drag forces
(Baker et al. 1983). However, explosively produced primary and secondary fragments are typically “chunky” or drag-controlled fragments. Thus, most methods
ignore the lift forces and evaluate the accelerations based on drag forces only
(U.S. Department of Defense 2008).
The equations that define the accelerations and velocities in the X and Y directions can be solved simultaneously to determine the distance traveled by the
fragment. Figure 8.8 represents a series of calculations for the resulting fragment
range given a variety of initial conditions. A number of initial trajectories were
228
FRAGMENTATION
10
(12.ρ0.CD.ADR)/M
5
1
0.5
0.1
0.05
0.01
0.01
1
10
(12.ρ0.CD.ADv02)/(M.g)
0.1
100
1,000
Figure 8.8 Fragment Range Prediction
considered to determine the maximum range; thus, the initial trajectory does not
have to be established to use Figure 8.8. Calculate the nondimensional velocity
[v], shown on the abscissa in Figure 8.8, with the following:
v=
12 · ρo · Cd · A D · vo2
M·g
(8.18)
where: ρ o = mass density of the medium through which the object travels (lbmsec2 /inch4 )
AD = drag area (inch2 )
g = gravity force (32.2 feet/sec2 )
From Figure 8.8 determine the dimensional range [R], shown on the ordinate.
Rearrange R to solve for the maximum range [R].
R=
M·R
12 · ρo · Cd · A D
(8.19)
8.5 FRAGMENT IMPACT DAMAGE
Almost any type of structure is susceptible to damage from explosively driven
fragmentation. Damage can range from cosmetic damage such as bent or cracked
architectural features to perforation (where a fragment passes through a wall
or structural element), penetration (where a fragment strikes and lodges in a
FRAGMENT IMPACT DAMAGE
229
structural element), or even global collapse (when a larger fragment strikes a
target with enough momentum to cause a primary element to fail).
Detonation of explosives can result in the formation of primary and secondary
fragmentation. These fragments can range from small primary fragments with
initial velocities in the order of thousands of feet per second to larger secondary
fragments with initial velocities of hundreds of feet per second. When a fragment
strikes a target, it will perforate the target, become embedded in the target (i.e.,
penetrate the target either with or without spalling), or be completely deflected by
the target. Factors that influence the fragment/target response include the initial
fragment velocity, the distance between the explosion and the target, the angle
at which the fragment strikes the target (angle of obliquity), and the physical
properties of the fragment (mass, shape, and material strength) and the target
(strength and thickness).
8.5.1 Fragment Penetration into Miscellaneous Materials
(THOR Equation)
Empirical THOR equations were developed based on testing for metallic and
nonmetallic materials (Ballistic Analysis Laboratory 1961, Ballistic Research
Laboratory 1963). All testing was performed to simulate primary fragments, and
the resulting data should be applied accordingly. The test data were assembled
into the THOR reports by Greenspon in 1976 (Greenspon 1976). These data
were gathered for military purposes and as such should be used in commercial
and industrial structures with specific limitations in mind. The method assumes
small fragment sizes into specific targets. The equation for minimum thickness
of a plate to resist perforation by a mild steel fragment is:
t=
where:
10 −
C1 +7C2 +3C3
C2
Af
·mf
C
− C3
2
· Vs
1−C5
C2
C4
· (cos φ) C2
(8.20)
t = fragment penetration (m)
mf = fragment mass (kg)
Af = presented area of fragment (m2 )
Vs = impact velocity (m/sec)
φ = impact angle relative to normal from tangent (rad)
C1 – C10 = empirical constants (see Table 8.6)
Other useful THOR equations for application to fragment impact on structures
are residual velocity and residual fragment mass equations. The THOR equation
for residual velocity is:
C
Vr = Vs − 10(C1 +7C2 +3C3 ) · t · A f 2 · m Cf 3 · (1/ cos φ)C4 · VsC5
where: Vr = residual velocity of fragment after perforation (m/sec)
(8.21)
230
1.044
1.112
0.203
0.328
2.534
0.674
2.590
1.310
−0.093
1.631
2.877
2.801
2.118
0.441
Lead
Lexan
Magnesium
Nylon, bonded
Nylon, unbonded
Plexiglass, as cast
Plexiglass, stretched
Steel, face hardened
Steel, hard homogeneous
Steel, mild homogeneous
Titanium
Tuballoy
1.021
0.583
1.103
0.889
0.889
0.674
0.835
1.144
1.092
0.720
0.499
0.705
3.431
0.678
1.046
Copper
1.029
1.042
Glass, bullet-resistant
0.314
Cast iron
C2
Doron
2.907
1.037
Aluminum (2024-T3)
C1
Target Material
Table 8.6 THOR Constants
1.369
1.073
−1.035
0.865
0.990
−0.654
−1.095
0.743
−0.968
−0.603
1.050
−1.117
1.262
0.773
−1.657
−0.945
0.655
−0.502
1.262
0.690
−0.723
−0.945
0.917
−1.014
0.989
0.846
−0.730
0.715
1.028
−1.051
−0.791
1.250
−1.072
−0.903
C4
C3
0.828
0.167
0.019
0.019
0.434
0.686
0.242
−0.162
0.392
−0.087
0.603
0.818
0.465
−0.362
0.802
0.523
−0.139
C5
−4.564
−1.927
−2.616
−3.017
−1.768
−6.431
−6.115
−6.619
−12.165
−6.026
−7.063
−3.267
−6.341
−10.161
−5.904
−9.052
−6.548
C6
0.560
1.086
0.138
0.346
0.234
0.437
1.402
−0.067
0.035
0.285
0.480
0.506
0.305
0.251
0.340
0.162
0.227
C7
0.447
−0.748
0.835
0.629
0.744
0.169
−0.137
0.903
0.775
0.803
0.465
0.350
0.429
0.343
0.568
0.673
0.694
C8
0.640
1.327
0.143
0.327
0.469
0.620
0.674
−0.351
0.045
−0.172
1.171
0.777
0.747
0.706
1.422
2.091
−0.361
C9
1.381
0.459
0.761
0.880
0.483
1.683
1.324
1.717
3.451
1.519
1.765
0.934
1.819
2.906
1.650
2.710
1.901
C10
231
FRAGMENT IMPACT DAMAGE
The THOR equation for residual fragment weight is:
C
m r = m f − 10(C6 +7C7 +3C8 −3) · t · A f 7 · m Cf 8 · (1/ cos φ)C9 · VsC10
(8.22)
where: mr = residual mass (kg)
The range for each variable in the THOR equations is given in Table 8.7.
8.5.2 Steel
Limit Velocity of “Chunky” Nondeforming Fragments for a Thin Metal Target
Fragment penetration into thin metal structural elements such as typical metal
buildings can be calculated based on the metal target material and fragment properties. A fragment velocity greater than the limit velocity (V 50 ) will penetrate the
metal target. The limit velocity for a thin metal target can be calculated using:
√
σt · ρt · V50n
(8.23)
V50 =
ρf
where:
σ t = target yield stress
ρ t = target density
ρ f = fragment density
V 50n = nondimensional limit velocity (calculate h/a and determine V 50n
from Figure 8.9)
h = target thickness
a = fragment radius
Fragment Penetration into Mild Steel Targets Fragment penetration calculations are based on some simplifying assumptions. The primary assumption is
that the fragment shape is a normal cylinder with a rounded leading edge (see
Figure 8.1). Additionally, the steel material is assumed to have a Brinell hardness of less than 150. Results will be conservative for materials with increased
hardness. Depth of penetration for mild steel targets impacted by armor-piercing
steel fragments is:
· Vs1.22
x = 0.30 · W 0.33
f
(8.24)
Depth of penetration for mild steel targets impacted by mild steel fragments is:
x = 0.21 · W 0.33
· Vs1.22
f
where:
x = depth of penetration (inches)
Wf = fragment weight (ounces)
V s = striking velocity (feet/sec)
(8.25)
232
Target Thickness
Range t mm
0.5 to 51.0
4.8 to 14.0
1.5 to 25.0
1.3 to 38.0
5.0 to 42.0
1.8 to 25.0
3.2 to 25.0
1.3 to 76.0
11.0 to 51.0
0.5 to 76.0
5.0 to 28.0
1.3 to 25.0
3.6 to 13.0
8.0 to 25.0
1.0 to 30.0
2.5 to 5.0
Target Material
Aluminum (2024-T 3)
Cast iron
Copper
Doron
Glass, bullet-resistant
Lead
Lexan
Magnesium alloy
Nylon, bonded
Nylon, unbonded
Plexiglass, as cast
Plexiglass, stretched
Steel, face hardened
Steel, homogeneous
Titanium alloy
Tuballoy
Table 8.7 Range of Variables in THOR Equations
1,372 to 3,078
213.4 to 3,170
182.9 to 3,658
762 to 2,987
152.4 to 3,353
61.0 to 2,897
91.4 to 3,048
304.8 to 3,658
152.4 to 3,200
304.8 to 3,505
152.4 to 3,170
61.0 to 3,048
152.4 to 3,353
335.4 to 3,475
335.0 to 1,859
365.8 to 3,353
Striking Velocity
Range Vs m/sec
1.9 × 10−3 to 3. 10 × 10−2 (30 to 475)
1.9 × 10−3 to 1.60 × 10−2 (30 to 240)
3.2 × 10−4 to 5.30 × 10−2 (5 to 825)
9.7 × 10−4 to 1.60 × l0−2 (1 5 to 240)
4.6 × 10−8 to 4. 4 × 10−6 (7.14 × 10−4 to 6.79 × 10−2)
4. 6 × 10−8 to 4.4 × 10 −6 (7.14 × 10−4 to 6.79 ×
10−2 )
4 .6 × 10−8 to 1.9 × 10−6 (7.14 × 10−4 to 2.96 × 10−2 )
4 .6 × 10−8 to 7.6 × 10−6 (7.14 × l 0−4 to 1.18 × 10−1 )
9.7 × 10−4 to 1.60 × l0−2 (15 to 240)
4.6 × 10−8 to 2.2 × 10−6 (7.14 × 10−4 to 3.43 × 10−2 )
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
1.4 × 10−7 to 4.4 × 10−6 (2.14 × 10−3 to 6.79 × 10−2 )
2.3 × 10−8 to 5.6 × 10−6 (3.57 × 10−4 to 8.57 × 10−2 )
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
3.2 × 10−4 to 1.60 × 10−2 (5 to 240)
Fragment Size Range mf kg (grains)
FRAGMENT IMPACT DAMAGE
233
12
10
ρf .V50
√ σt . ρt
8
6
4
2
0
0.2
0.4
0.6
0.8
1.0
1.2
h/a
1.4
1.6
1.8
2.0
2.2
Figure 8.9 Nondimensional Limit Velocity versus Nondimensional Thickness for
“Chunky” Nondeforming Fragments (U.S. Department of Energy 1992)
Most fragments generated during an explosion will not resemble a simplified
bullet—the majority of primary fragments are more blunt; thus, the bullet-shaped
fragment is conservative. Certainly some sharper-edged fragments are generated;
however, the likelihood that these fragments will strike the target in the appropriate orientation to control the design is minimal.
8.5.3 Fragment Penetration into Concrete Targets
When a fragment strikes a concrete target, the impulse imparted by the fragment
on the concrete will cause a crater of irregular size. The depth of the crater grows
with impact velocity for a given fragment size (or with a reduction in fragment
cross-sectional area for a given velocity). If the velocity is high enough (typically
greater than 1,000 feet/second [U.S. Department of Defense 2008]), the fragment
will penetrate beyond the crater. If the fragment deforms upon impact, the crater
size will decrease or may not form at all.
In addition to the crater on the impact side, a crater may form (generating concrete spall) on the opposite side of the target adjacent to impact. This crater is the
result of a compression wave passing through the concrete material and reflecting
off the free surface of the opposite concrete face. The reflection causes tension
stresses at the surface of the concrete element. If the tension stress exceeds the
compressive strength of the concrete, spalling of the concrete target will occur.
If the impulse imparted on the target is sufficient, the spall crater may extend to
the reinforcing steel. As impact velocity increases, the impact and spall craters
increase, and the fragment will become lodged in the target or pass through the
target (perforate).
234
FRAGMENTATION
Armor-Piercing Fragment Penetration into Concrete Targets The maximum
penetration of an armor-piercing fragment into a massive concrete target is defined as:
For Xf ≤ 2·d:
X f = 4.0 · 10−3 ·
√
K · N · D · d 1.1 · Vs0.9
(8.26)
For Xf > 2·d:
X f = 4.0 · 10−6 · (K · N · D) · d 1.2 · Vs1.8 + d
(8.27)
where: Xf = maximum penetration by armor-piercing
fragment (inch)
K = penetration constant = K = 12.91
f c
N = nose shape factor (see Figure 8.1)
D = caliber density (see Figure 8.1)
d = fragment diameter (inch)
For a standard primary fragment and concrete strength (f c ) of 4,000 psi, Equations 8.26 and 8.27 can be expressed in terms of fragment diameter (inches) as:
For Xf ≤ 2·d:
X f = 2.86 · 10−3 · d 1.1 · Vs0.9
(8.28)
X f = 2.04 · 10−6 · d 1.2 · Vs1.8 + d
(8.29)
For Xf > 2·d:
Or in terms of fragment weight:
For Xf ≤ 2·d:
X f = 1.92 · 10−3 · W 0.37
· Vs0.9
f
(8.30)
1.8
+ 0.695 · W 0.33
X f = 1.32 · 10−6 · W 0.4
f · Vs
f
(8.31)
For Xf > 2·d:
For concrete strengths other than 4,000 psi, the penetration depth can be estimated by multiplying the results by the square root of the ratio of the concrete
strengths:
X f = X f ·
4000
f c
(8.32)
where: X f = maximum penetration by armor-piercing fragment into concrete
with compressive strength of f c (inch)
FRAGMENT IMPACT DAMAGE
Table 8.8
235
Fragment Penetration Factors, K3
Type of Material
K3
Armorpiercing steel
1.00
Mild steel
0.70
Lead
0.50
Aluminum
0.15
Non-Armor-Piercing Fragment Penetration into Concrete Targets For nonarmor-piercing fragments, the depth of penetration into a concrete target can be
calculated by taking the fragment metal hardness into account:
X f = K3 · X f
(8.33)
where: X f = maximum penetration by non-armor-piercing fragment (inch)
K 3 = constant to account for fragment metal hardness (see Table 8.8)
8.5.4 Fragment Perforation of Concrete Targets
The equations presented in Section 8.5.3 assume the concrete is of infinite thickness. Of course the material thickness of a concrete slab of finite thickness required to prevent perforation is greater than the maximum penetration depth
into an infinite thickness (due to confinement and compressive resistance of a
massive concrete structure). The minimum concrete thickness required to prevent perforation as a function of maximum penetration into a massive concrete
structure is:
T p f = 1.13 · X f · d 0.1 + 1.311 · d
(8.34)
where: Tpf = minimum concrete thickness to prevent perforation by design fragment (inch)
Xf = maximum fragment penetration into a massive concrete structure
as described in Section 8.5.3
If a fragment perforates a concrete target, it can pose a threat to personnel
and structural elements behind the concrete target. The residual velocity of the
fragment can be calculated using equations that define the fragment velocity as
a function of penetration depth:
For Xf ≤ 2·d:
0.555
Tc 2
Vr
= 1−·
Vs
Tp f
(8.35)
236
FRAGMENTATION
For Xf > 2·d:
0.555
Tc
Vr
= 1−·
Vs
Tp f
(8.36)
where: Tc = concrete thickness up to Tpf (inch)
Vr = residual velocity of fragment as it leaves the element (feet/sec)
8.5.5 Fragment Spalling of Concrete Targets
As discussed in Section 8.5.3, concrete spall occurs when the impacting fragment
imparts enough impulse to generate a compression wave and a resulting tension
wave in the concrete that exceeds the compressive strength of the concrete. The
spall crater typically does not exceed the reinforcement depth. Concrete spalling
is a function of fragment penetration depth. Therefore, the minimum concrete
thickness required to prevent spalling as a function of maximum penetration into
a massive concrete structure can be expressed as:
Tsp = 1.215 · X f · d 0.1 + 2.12 · d
(8.37)
where: Tsp = minimum concrete thickness to prevent spall by design fragment
(inch)
Xf = maximum fragment penetration into a massive concrete structure
as described in Section 8.5.3 (inch)
If spalling occurs, it is likely the concrete spall will generate a secondary fragment that is hazardous to personnel and equipment behind the concrete target.
While the velocity produced when the concrete material fails is likely to be relatively low, the spall fragment velocity will also be affected by the wall motion
in response to the blast loading. Therefore, it is typically accepted that concrete
spall should be prevented by designing the concrete thickness to be sufficient to
avoid spall, or by providing an additional cover for the concrete surface (i.e., a
catch system or fiber-reinforced laminate).
8.5.6 Roofing Materials
Most roofing materials are relatively lightweight and have low thresholds of serious damage. However, the angle of obliquity will likely reduce the effect of fragment damage. Lower limits for fragment damage to miscellaneous lightweight
roofing materials based on fragment momentum are shown in Table 8.9.
REFERENCES
237
Table 8.9 Fragment Impact Damage for Roofing Materials (Baker et al. 1977)
Roofing Material
Minimum Fragment
Momentum (mV) for
Serious Damage
(lbf-sec)
Comments
Asphalt shingles
0.159
CracksShingle
1.370
Damage deck
<0.159
Crack tar flood coat
0.451
Crack surface of conventional built-up
roof (BUR) without top layer of stones
>0.991
Crack surface of conventional BUR with
2.867 lb/feet2 (137 Pa) top layer of
stones
Built-up roof
Miscellaneous
1/8-inch asbestos cement
shingles
0.159
1/4-inch asbestos cement
shingles
0.285
1/4-inch green slate
0.285
1/4-inch gray slate
0.159
1.2-inch cedar shingles
0.159
3/4-inch red clay tile
0.285
Standing seam terne metal
0.991
Plywood deck cracked
8.5.7 Other Materials
A number of other materials have been evaluated for fragment penetration, including sand, various rock and other geological materials, and combinations of
sand and concrete panels. While they are not presented here, they are covered in
varying detail in Structures to Resist the Effects of Accidental Explosions (U.S.
Department of Defense 2008), Explosion Hazards and Evaluation (Baker et al.
1983), and A Manual for the Prediction of Blast and Fragment Loadings on
Structures (Department of Energy 1992), as well as the Department of the Army
Historical Summary (DAHS) manual Fundamentals of Protective Structure Design for Conventional Weapons (Department of the Army 1998).
REFERENCES
Baker W. E., P. A. Cox, P. S. Westine, J. J. Kulesz, and R. A. Strehlow. 1983. Explosion
Hazards and Evaluation. Fundamental Studies in Engineering 5. Amsterdam: Elsevier
Scientific Publishing Company.
Baker W. E., J. J. Kulesz, R. E. Ricker, R. L. Bessey, P. S. Westine, v. b. Parr, and g. a.
Oldham. 1977. Workbook for Predicting Pressure Wave and Fragmentation Effects
238
FRAGMENTATION
of Exploding Propellant Tanks and Gas Storage Vessels, NASA CR-134906, NASA
Lewis Research Center.
Ballistic Analysis Laboratory, Johns Hopkins University. 1961. The Resistance of Various Metallic Materials to Perforation by Steel Fragments; Empirical Relationships for
Fragments Residual Velocity and Residual Weight (Project THOR Technical Report
47). Baltimore, MD: Johns Hopkins University, Institute for Cooperative Research.
Ballistic Research Laboratory. 1963. The Resistance of Various Non-Metallic Materials to Perforation by Steel Fragments, Empirical Relationships for Fragment Residual
Velocity and Residual Weight (TR 51). Aberdeen Proving Ground, MD: Ballistic Research Laboratory.
Department of the Army. 1998. Fundamentals of Protective Structure Design for Conventional Weapons (TM 5-855-1, Air Force AFPAM 32-1147[I], Navy P-1080, DSWA
DAHSCWEMAN-97). For Official Use Only. Washington, DC: Department of the
Army.
Greenspon J. E. 1976. An Approximate Non-Dimensional Representation of the THOR
Equations (U.S. Army Materiel Systems Analysis Activity TR #173).
Gurney R. W. 1947. The Initial Velocities of Fragments from Bombs, Shells, and Grenades
(Report No. 648). Aberdeen Proving Ground, MD: Ballistic Research Laboratory.
Mott R. I. 1943. A Theoretical Formula for the Distribution of Weights of Fragments
(Ministry of Supply, AC-3642).
U.S. Department of Defense (DoD). 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense.
U.S. Department of Energy. 1992. A Manual for the Prediction of Blast and Fragment
Loadings on Structures (DOE/TIC-11268). San Antonio, TX: Southwest Research Institute.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
III
System Analysis and Design
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
9
Structural Systems Design
Robert Smilowitz and Darren Tennant
9.1 GENERAL DISCUSSION
In order to protect the occupants of a building that is either the target of an explosive attack or the unintended victim of collateral damage resulting from an attack
against a neighboring property, attention must be paid to the performance of the
structural system. Although the structure may require extensive repair following
an incident, the primary goal of protective design is to prevent structural collapse
and to minimize debris. This goal is accomplished through the design and detailing of structural components to sustain specified intensities of blast loading, and
the design and detailing of structural systems to provide ductile and redundant
load paths, regardless of the cause or event. This combination of threat-specific
and threat-independent design provides the most robust structural system that is
able to survive both the anticipated and the unimaginable catastrophic events. As
with the performance-based design of structures to resist strong ground motions,
the design of structures to resist blast loads considers successive levels of damage and deformation in response to increasingly severe events. Also similar to
seismic design, the analytical methods that are required to demonstrate the performance of structural components and structural systems in response to extreme
blast loading are necessarily dynamic and inelastic.
9.1.1 Seismic versus Blast
Performance-based methods are used to design structures that must resist strong
ground motions and structures that must resist blast loading. In both cases, different extents of acceptable damage are specified for events of different intensity. This approach acknowledges the financial burden associated with comprehensive protection in response to the most extreme loading conditions and rationally accepts successively greater risk for successively more severe events.
This risk-based approach starts with a risk assessment that considers the function, criticality, occupancy, site conditions, and design features. However, unlike
natural hazards, man-made events are entirely unpredicable. As a result, risk
associated with blast loading can at best be quantified in a relative sense, in order to compare different levels of protection or different protective measures.
241
242
STRUCTURAL SYSTEMS DESIGN
Since there are not enough data to justify the absolute quantification of risk, the
choice of design basis threat is inherently subjective and is often based on past
experience.
The analytical methods that are used to evaluate the performance of structures in response to these forms of extreme loading involve the solution of the
equations of motion. Since ground motions are typically assumed to be applied
uniformly over the foundations of most structures, the spectral approach allows
the designer to envelop the expected response without having to consider a specific input motion time history. However, blast loads and blast effects are typically localized to the immediate vicinity of the detonation, and these enveloping
methods are not applicable. Localized blast loads will deform the exterior bays
of the structure to a much greater extent than the interior bays, and, in the time
frame of the applied loading, the story shears may not necessarily be distributed
through the diaphragms in proportion to the framing stiffness. Typically, the ductility demands at the blast-loaded face of the structure far exceed the demands
elsewhere throughout the remainder of the structure.
Some types of construction are inherently more ductile than others, and these
are better suited for seismic and blast-resistant design. While cast-in-place concrete can be detailed to provide sufficient ductility, post-tensioned two-way flatslab structures are problematic. Punching shear failures at the slab joint must be
avoided under lateral sway deformations, and these systems possess limited lateral deformation capability. Although a two-way flat-slab system might perform
satisfactorily in response to strong ground motion when it is used in combination with a stiffer shear wall system that controls the ultimate lateral sway of
the structure and prevents destabilizing P-Delta effects, the punching shear failure vulnerability is a function of the slab performance in response to blast infill
pressures and is not improved with the addition of shear walls.
The fundamental difference between seismic excitations and blast-resistant
design is the distribution of ductility demand throughout the structure. Strong
ground motions engage the entire lateral resisting system, while resistance
to blast loading is more concentrated in the structure in the vicinity of the
explosion. Moment frames may provide high ductility and energy absorption
when the plastic hinging of beams is spread throughout the structure. In
order for inelastic deformation to be distributed throughout the structure, the
beam-column connections must develop the flexural hinge in the beam. The
elements must also be proportioned to prevent plastic story mechanisms that
might limit the inelastic energy absorption to isolated regions of the frame. This
is achieved through the “strong-column, weak-beam” (SCWB) approaches for
both seismic and blast design. Since blast loading is more localized than seismic
excitations, the resulting plastic hinging is more localized, with more severe
rotations than those that result from strong ground motions. Structural systems
that effectively resist the lateral forces resulting from strong ground motions,
such as concentrically braced frames (CBF) and eccentrically braced frames
(EBF), must be considered on a case-by-case basis for use in blast-resistant
design.
GENERAL DISCUSSION
243
However, the manner in which seismic and blast forces are distributed
throughout the building is strongly affected by the size, shape, and location of
structural elements. Although seismic forces are proportional to the mass of the
building, inertial resistance reduces the response to blast loading. Structures that
are designed to resist both blast forces and strong ground motion benefit from
low height-to-base ratios, balanced resistance, symmetrical plans, uniform sections and elevations, the placement of shear walls and lateral bracing to maximize
torsional resistance, short spans, direct load paths, and uniform floor heights.
Although the benefits of employing these design elements may be similar, the
reasons for incorporating these design features may differ. For example, seismic
excitations may induce torsional response modes in structures with reentrant corners, whereas these conditions provide pockets where blast pressures may reflect
off adjacent walls and amplify the blast effects. Similarly, first-floor arcades that
produce overhangs or reentrant corners create localized concentrations of blast
pressure and expose areas of the floor slab to uplift blast pressures. Adjacent
structures may impact one another as they respond to base motions, while adjacent structures may be subjected to multiple reflections of blast waves. The
geology of the site influences the characteristics of the ground motion that load
the structure, and the geology influences the size of the blast crater and the reflectivity of the blast waves off the ground surface.
The ductility demands for both seismic and blast-resistant design require attention to details. Confinement of the concrete core prevents premature buckling
of the rebar and maintains load-carrying capacity, despite the significant cracking associated with large inelastic deformation. Closely spaced ties and spiral
reinforcement are particularly effective in increasing the ductility of a concrete
compression element. Except for near-contact detonations, carbon fiber wraps
and steel jacket retrofits provide comparable confinement to existing structures;
however, carbon fiber materials may be damaged when the standoff distances
are too short. Steel column splices must be located away from regions of plastic hinging or must be detailed to develop the full moment capacity of the section. Closely spaced stiffeners or concrete encasement will prevent local flange
buckling of steel sections. Reinforced concrete beam sections require equal resistance to positive and negative bending moments. In addition to the effects of
load reversals and rebound, doubly reinforced sections possess greater ductility
than singly reinforced counterparts. Steel beams may be constructed composite
with the concrete deck in order to increase the ultimate capacity of the section;
however, this increase is not equally effective for both positive and negative moments. While the composite slab may brace the top flange of the steel section, the
bottom flange is vulnerable to buckling. Tube sections and concrete encasement
are particularly effective in preventing flange buckling under load reversals.
9.1.2 Analytical Methods
A wide variety of analytical approaches may be used to determine the performance of structures in response to explosive loading. The various methods
244
STRUCTURAL SYSTEMS DESIGN
include empirical explosive test data that can be expressed as prescriptive
requirements or pressure-impulse charts, single-element response analysis,
structural system response analysis, and detailed finite element analyses. These
dynamic response methods all characterize the behavior of the structural
system in response to a short-duration but high-intensity dynamic loading.
The dynamic analysis methods must be able to capture the frequency of the
loading function and the dominant frequencies of the structural system, and
represent the governing failure mechanisms so as not to overlook critical failure
mechanisms that may precipitate collapse or hazardous debris. The designer
must therefore anticipate the likely failure mechanism before selecting the
appropriate analytical approach.
9.2 MODELING
Long before complex finite element analysis methods were readily available, the
design of structures to resist blast loading relied extensively on single-degree-offreedom (SDOF) methods. These methods were generally conservative as long as
the designer anticipated the system response and made sure the analytical model
captured all failure mechanisms and response characteristics. Advanced finite element methods may be used where structural complexity or spatial distribution of
loading requires more detailed analyses. These methods can represent complex
geometries, connectivity, and material properties. While simplified models are
prone to problems associated with too little information, complex models need
to be tailored to available resources and time constraints. The optimal model
therefore contains enough information to capture the response of interest without becoming so excessively complex that it cannot solve the equations of motion within a reasonable solution time. Accepted practice requires that enough
resolution be included in the model to accurately capture response modes.
When fast-running simplified methods are deemed inadequate, detailed
analytical models can often be constructed to calculate structural performance in response to blast and impact. The current state of the art relies on
explicit dynamic inelastic finite element software such as NLFlex, LS-DYNA,
AUTODYN, Abaqus, Pronto, and the like. When modeling a significant section
of a building, such as a couple of bays, the required level of resolution or
element size to calculate an accurate response on current-generation computer
hardware limits the extent of the structure that can be analyzed. For these 3D
models, the required computer RAM increases by the cube of the resolution,
and the computational time increases by the 4th power. As a result, the analysis
of larger models will become feasible as computer hardware advances.
At this resolution level, each rebar is typically represented explicitly with
beam or bar elements, and structural steel is typically represented explicitly with
shell or plate elements. Concrete, soil, and rock are represented with continuum
elements. In view of this background, macroscopic phenomenological constitutive models are the state of the art for common construction materials such
MODELING
245
as steel and concrete. Research efforts on more fundamental meso- and micromechanical approaches fill the literature; however, these are currently too computationally demanding for production use.
The goal of a phenomenological model is to accurately capture the important
characteristics of material response in the expected loading regime, while obeying the laws of thermodynamics. The theory must be implemented in a robust
algorithm that is efficient in terms of both RAM and CPU usage. Important material characteristics for concrete include strength dependence on pressure, energy absorption or hysteresis due to pore compaction, increase of resistance with
loading rate, and softening due to fracture or crushing. Concrete also transitions
from a brittle material that fractures at low pressures to a more ductile material
that flows at higher pressures. Important considerations for steel include hardening after yield, strength dependence on loading rate, and fracture dependence on
confining pressure. A realistic representation of fracture or material softening is
very important in blast applications. Once softening initiates, material damage
localizes into discrete cracks or narrow bands. Maintaining accurate simulations
during localization requires some sort of regularization. At currently achievable
resolution levels, this usually requires a fracture energy-based approach whereby
the softening stress-strain response is a function of element dimension.
Theories of plasticity, viscoplasticity, and/or continuum damage mechanics
are a few of the popular theoretically sound frameworks for phenomenological
models. A well-formulated model will use parameter identification from a small
number of lab tests to reasonably characterize material response over a very wide
range of loading conditions.
9.2.1 Systems
Several of the most common structural systems will be discussed, including the
significant features that must be represented in the analytical models. While this
list represents a sampling of structural systems that may be analyzed, it is not intended to limit the types of structural systems that may be used for blast-resistant
design.
Concrete and Steel Moment Frames Moment frame buildings may be analyzed one component at a time, modeling each component as an SDOF system
with dynamic blast loads either applied tributary to the loaded surface area or collected through the reaction forces from secondary members as they frame into
the primary structure. The appropriate performance limits for the calculated inelastic deformations are typically conservatively specified, particularly for axial
members, where P-Delta effects may initiate structural instability. Furthermore,
simplified SDOF methods are incapable of capturing localized steel flange deformations, steel splice performance, concrete panel zone deformations, and other
localized response characteristics. These SDOF approaches are only appropriate for standoff detonations that apply fairly uniform loads to the member. The
overall frame response may similarly be determined using an SDOF model of
246
STRUCTURAL SYSTEMS DESIGN
the building’s lateral resisting system. These blast-induced base shears recognize
the flexibility of the frame and a characteristic extent of frame ductility; however,
since blast damage is typically localized in proximity to the point of detonation,
these overall ductility levels are approximate and represent overall energy dissipation of the framing system. Fundamental to the modeling and analysis of
the frame using these methods is the appropriate detailing of the connections so
the structural members will be capable of developing the calculated extent of
rotation and lateral drift.
Steel-Braced Frames and Concrete Frames with Concrete Shear Walls Similar to the response of concrete and steel moment frames, the response of steelbraced frames or concrete frames with concrete shear walls may be analyzed one
component at a time, to determine local response in proximity to the detonation;
however, the global response of the structure will be characterized by a stiffer
lateral resisting system with a higher fundamental frequency of response. Not
only will the period of the structural system be shorter than that of a comparable
moment frame structure, but the amount of energy dissipation will be less. This
type of structural system will typically develop larger blast-induced base shears,
and the floor diaphragms must be capable of distributing the collected lateral
loads to the various lateral resisting components.
Precast Tilt-Up with Concrete Shear Walls Precast tilt-up load-bearing walls
may be modeled as single-degree-of-freedom systems, provided the detonation
is sufficiently far from the building to produce a uniform loading over the floor
height. The flexural deformations of these wall systems must be limited in order
to prevent P-Delta effects that may precipitate instability, and the diaphragms
must be sufficiently tied to the wall in order to transfer the large lateral loads and
rebound forces. Subsequent diaphragm models are required in order to demonstrate their ability to distribute loads to the concrete shear walls.
Reinforced Masonry Walls with Rigid Diaphragms Reinforced masonry
walls may be analyzed as single-degree-of-freedom systems; however, pressureimpulse diagrams that represent empirical data and the results of analytical studies may be used to evaluate their performance in response to blast loading. As
with precast tilt-up construction, analyses must be performed in order to demonstrate that the diaphragms are capable of distributing the lateral loads to the shear
walls and are able to restrain the large rebound forces.
9.2.2 Materials
Structural materials are affected by extremely high loading rates, high confining pressures, and large inelastic deformations in response to blast loading.
Simplified elastic-plastic resistance functions are often used to represent
ductile materials with dynamic enhancement factors that account for the strain
MODELING
247
rate loading. Explicit dynamic inelastic finite element analyses require more
sophisticated material models to represent the behavior of structural material.
Materials such as concrete and steel are typically stronger than the minimum
values specified on the construction documents. Although ASTM specifications
and codes define the minimum properties for materials, the actual materials being
installed typically have strengths that exceed these minimum requirements. For
the analysis and design of reinforced concrete and structural steel elements for
flexure, it is generally acceptable to increase the steel material’s minimum yield
strength by 10%. Ignoring the material’s average strength and dynamic strength
increase generally results in increased factors of safety for bending elements;
however, underestimating the effects of material overstrength is nonconservative
for shear and connection design.
Structural elements also exhibit higher strengths when subjected to dynamic
blast loads than they exhibit in response to static loads. Material testing demonstrates an increase in strength in response to high-strain-rate dynamic loadings.
As with the use of average strength factors, underestimating the high-strainrate effects produces greater factors of safety for flexural element design but
is nonconservative for shear and connection design. However, the factored yield
strength that will be used for determining the flexural resistance function should
not exceed the ultimate dynamic strength of the material.
Threat-Dependent Effects Explosive threat can be grouped into three general
categories based on the explosive weight and standoff, contact/near contact, and
near field and far field. Contact and near contact place the explosive threat in
close proximity to the structure, generally at scaled ranges less than 0.5, scaled
standoff range (Z = R/W 1/3 ), where R is standoff, and W is explosive weight. At
this standoff distance, the explosive effects are highly localized, and the air-blast
loads have extremely high-pressure short-duration pulses. At these scaled ranges,
use of simplified tools is questionable. Scaled ranges between 0.5 and 2 fall into
the near-field range. If charge shape is not an issue and the structure surface is
flat and regular, simplified air-blast loading functions are applicable. Far-field
threats generally exist at scaled ranges greater than 2.0. Pressure loading is more
uniform across the structure with lower peak pressures and long duration pulses.
Air-blast clearing can be a significant issue for far-field threats.
Near Range versus Far Range The amplitude of peak pressures for explosive
detonations that are in close proximity to a structure or structural element is extremely high. At these short distances the blast loading is taking place within
the explosive by-products, and the reflection factors may be in excess of 10. Although the duration of the applied loads is short, particularly where complete
venting of the explosion occurs, the corresponding impulses are also extremely
high. Fragments associated with the high-pressure range usually consist of highvelocity primary fragments from the explosive casing breakup or the acceleration
of objects positioned close to the explosion. Simplified analytical tools that are
based on the modified Friedlander equations (Hyde 1992, Kingery and Bulmash
248
STRUCTURAL SYSTEMS DESIGN
1984) become less reliable for scaled ranges that are less than 0.37 to 0.45. Advanced analysis and testing show that when charge shapes other than spheres or
hemispheres are considered, simplified methods lose accuracy at scaled ranges
as high as 2.0. Advanced analytical methods based on computational fluid dynamics (CFD) are required to calculate accurate loading for near contact or for
complicated surfaces.
For near contact, the durations of the applied loads are much shorter than the
time it takes the structural elements to reach maximum deflections; they can often be analyzed for the impulse rather than the peak pressure associated with
longer-duration blast pressures. Also associated with the “close-in” effects of a
high-pressure design range is the possible occurrence of spalling and scabbing
of concrete elements and post-failure fragments. Spalling and scabbing result
from the disengagement of the concrete cover over the reinforcement on the
back side of a concrete structure (the side away from the explosion). Spalling
is caused by the high-intensity-blast shock wave reflecting off the back surface as a tension wave, and scabbing is associated with the disengagement of
concrete as it undergoes large displacements. Secondary fragments that result
from spalling and scabbing of structural materials can be a hazard to people and
equipment.
Structures that are vulnerable to close-in detonations are vulnerable to breach
and direct shear. Protective and containment structures adjacent to occupied areas should be constructed with double walls separated by a void of sufficient size
to allow for the deflection of the outer sacrificial wall. The relative strength of the
two walls and the required space between them may be determined on a case-bycase basis. Although simplified approximations may provide design guidance,
explicit dynamic nonlinear finite element analyses are required to accurately determine the potential for breach and spall.
Where scaled ranges are large, the duration of the blast loading may exceed
the response time of the structure. Dynamic or quasi-static response characteristics may be observed for this type of loading. Simplified single-degree-offreedom methods may accurately determine flexure and flexural shear failures
that may result from far-range loading.
9.2.3 Members
This section will discuss the modeling of structural components, such as
columns, beams, slabs, and walls, to determine their behavior in response to blast
loading. Included in this discussion will be the evaluation of deformation and reaction forces associated with flexure, shear, and axial effects. Also discussed will
be issues such as breach, spall, and fracture.
In a numerical model, the more accurate the representation of a member, the
better the results will be. Modeling of structural components, such as columns,
beams, slabs, and walls, can be done on the component level, as an assembly or as
a complete structure. When they are modeled on the component level, care must
be taken to prevent the results from being adversely impacted by the assumed
MODELING
249
boundary conditions. Generally boundary conditions will be either free, fixed,
pinned, or a combination. It is good practice to vary the boundary conditions to
see the impact each choice has on the response of the component. Fixing the
axial response at both ends of a member will result in a stiffer section than a
member with a fixed and a free end. Application of boundary conditions directly
to the ends of a member can result in additional confinement to pressure-sensitive
materials, such as concrete. This additional confinement can prevent shear failure
and force a member into a flexural response. Single-component models allow
the analyst the opportunity to perform grid refinement studies to determine the
impact of resolution on the accuracy of the solution. While single-component
models are the simplest to construct and the fastest to run, the results may not
provide the required accuracy.
The next step up from a single-component model is an assembly model. Examples would be a column with the addition of bracing beams, or a wall with
beams and columns surrounding. The increased model size and run time for
assembly models can improve accuracy due to more realistic boundary representation. The boundary conditions are further removed from the component of
interest, resulting in a better representation of the physical system. The increased
complexity of assembly models necessitates the introduction of connection details. This is especially true in modeling of steel-frame structures.
Full-structure models provide the greatest accuracy in modeling of a structure
if the level of refinement can be maintained. The trade-offs among model size,
run time, and level of refinement usually result in models of a full structure having a lower level of refinement. The reduction in refinement is traded off versus
the more accurate boundary conditions and the addition of the overall response
and flexibility of the full structure.
Advanced finite element methods using accurate material models are capable
of capturing the response of members through failure. Level of damage can be
obtained directly from the analysis results by evaluating deformations, velocities,
strains levels, and material damage metrics. Simplified analysis approaches rely
on guidelines based on comparison with testing or advanced analysis to relate
displacements or rotations with level of damage. These guidelines, which can
be found in military manuals and design references, are generally conservative
(Unified Facilities Criteria 2002, Unified Facilities Criteria 2008b). This will
better enable a structure to survive the applied load. The available test databases
are not extensive enough to cover all possible design configurations; as a result,
the accuracy of predicted damage can be affected.
Shear and Flexural Reaction Forces For most structural members, their response when subjected to a blast loading is a combination of shear and flexure.
How the member responds is a function of the structural component’s properties and the intensity of the loading. Long, slender members are more likely to
respond in flexure than short, thick sections. The intensity of the load also impacts the type of response. A more impulsive loading is more likely to produce
a shear response, as compared to a longer-duration load. As a result, a column
250
STRUCTURAL SYSTEMS DESIGN
that responds primarily in bending due to an explosive threat at a large standoff
distance may respond in shear due to a near-contact blast.
Reaction forces collected at the boundaries of component models can be extreme. Reactions from shear deformation modes are usually of short duration
with a high peak force. Reactions from primarily bending modes have a longer
duration on the order of the period of the member. Whether a section deforms in
shear or flexure depends on the properties of the section and the applied load.
Inclusion of supporting members or modeling of a region extending beyond
the component of interest will result in more accurate representation of reaction
forces. When the compliance of bracing members is included in the modeling,
the reaction at the column-beam intersection will be reduced. This allows for a
more efficient and cost-effective design of a connection.
Axial Effects Accounting for large deformation effects such as arching, stress
stiffening, tension membrane, and P-Delta buckling is critical to the accurate
evaluation of structures subjected to blast. Simplified modeling approaches, such
as SDOF, do not account for large deformation effects in their resistance functions. As a result, the prediction of response is usually based on rotations or
displacements, and the level of damage is evaluated based on guidance from calibration of the models to test results or an advanced analysis database. Extrapolation outside of the available test database can result in inaccurate assessments of
structural damage. Advanced finite element methods are formulated to account
for geometric nonlinear and large deformation effects and are better suited for
evaluating response over the full range of damage through collapse.
The additional confinement provided by axial preload can influence the
response of concrete structural components such as columns or load-bearing
walls subjected to blast loading. Strength dependence on confining pressure
can increase the capacity of concrete materials subjected to blast loading. In
general a vertical structural component loaded with a lateral blast pressure will
initially experience a compression membrane, followed by a tension membrane
as the deformations increase. During compression membrane behavior, the
structural resistance to lateral deformation may increase. However, as the section
deforms due to blast loading and the lateral deformation grows, the ability to
resist deformation transitions from compression arching to tension resistance.
Post-blast damage stability is dependent on the damaged section’s ability to
continue to carry axial load. Geometric P-Delta effects can result in collapse
under axial loads.
The response of steel columns subjected to blast is less sensitive to initial axial
load. The addition of an axial preload does have an impact on the post-damage
failure of the section or collapse due to P-Delta effects. A common method for
evaluating the response of axially loaded members is to break the response up
into load phases. Initially the member is preloaded with the axial force. The top
axial degree of freedom is fixed, locking in the preload. The member is then
subjected to the lateral blast load, and the damage to the column due to blast is
evaluated. The axial degree of freedom is then released to allow either the axial
MODELING
251
force on the column to stabilize or the damaged column to collapse under the
axial load.
Fracture Capturing fracture of a member requires the use of high-fidelity
modeling. Simplified methods such as SDOF models can be calibrated to
testing or high-fidelity modeling results to give indications of when fracture
occurs. Displacement or rotation limits can be used to predict failure, but these
are usually conservative. Advanced finite element methods, using constitutive
models that capture the response of materials undergoing large strain and
softening, are required to investigate fracture of members. The evaluation of
breach and spall is impacted by the model’s ability to calculate fracture in the
material. If fracture is not accurately modeled, the ability to model breach and
spall will be compromised.
9.2.4 Connections
This section will discuss the modeling of structural connections in order to determine their behavior and performance in response to blast loading. Included in
this section will be the effects of shear and flexure associated with small deformations and the axial forces generated by large deformations.
Modeling of connections is critical for accurately predicting the response of
structures subjected to blast and the potential for progressive collapse following
a blast. Evaluation of connections is best accomplished using advanced modeling
approaches. The complicated nature of connections requires a refined treatment
of the details of their construction. For concrete structures, the additional confinement provided by the reinforcing steel in the joint can increase capacity and
ductility. Models of a concrete joint need to include the detailed geometry and
accurate material models that can account for the strength increase resulting from
this confinement. Steel connections can be complicated and require the modeling
of contact surfaces between members and connection plates and angles. Failure
of welds or bolts can dominate the response of joints, and detailed models are
required to address possible failure modes including fracture of welds or surrounding steel, bolt failure, and bearing failure of bolt holes.
Large Deformations The response of a connection changes as it deforms. Rotations can become large, and the resulting forces at the joint change. Initially at
small deformations the forces in a joint can be dominated by shear and compression. As the joint deforms more, the method of force transfer through the joint
can change to tension. The forces in the connection are impacted by a number
of effects, such as prying, twisting, fracture, failure in bearing, and local buckling. These complicated geometric and material effects are difficult to capture
accurately. Details of connection construction are difficult for simplified tools to
model accurately.
Shear versus Flexural Effects During blast loading, a connection can experience direct blast loads and loads transferred from neighboring members. As
252
STRUCTURAL SYSTEMS DESIGN
adjacent members fail, additional forces can be transferred through connections
from surrounding members. These forces can be transferred by shear and flexure. Bending of neighboring members can result in lateral forces due to tension
or compression of the member. Prying at the joint can also induce lateral loads.
9.3 ANALYTICAL APPROACHES
The analytical method that may be used to evaluate a structure’s response to
short-duration but high-intensity dynamic loading depends on the characteristics
of the loading and the anticipated failure mechanism. These approaches range
from simplified table lookups to multi-degree-of freedom finite element calculations, and each approach has inherent constraints or limiting assumptions that
must be appropriate for a specific application. For example, blast-resistant glass
thickness may be selected using ASTM F2248, Standard Practice for Specifying
an Equivalent 3-Second Duration Design Loading for Blast Resistant Glazing
Fabricated with Laminated Glass (ASTM 2003), and ASTM E1300, Standard
Practice for Determining Load Resistance of Glass in Buildings (ASTM 2007).
Empirical explosive test data can often be expressed as prescriptive requirements
or pressure-impulse charts. Flexural response behavior of structural components
can often be determined using single-degree-of freedom methods. The Component Explosive Damage Assessment Workbook (CEDAW) (Oswald 2005) and
Single-Degree-of-Freedom Blast Effects Design Spreadsheets (SBEDS) (Protective Design Center 2006) were developed by the U.S. Army Corps of Engineers
(ACE) for use in blast effects assessments and utilize the pressure-impulse (P-I)
and single-degree-of-freedom (SDOF) response methodologies. Detailed finite
element analyses are computationally intensive and require complex modeling
of the structure and material properties. The choice of the most efficient analytical approach requires a full understanding of the benefits and limitations.
In all cases, the analytical methods must be numerically robust in order to
characterize the transient blast loading pulse and the dominant frequencies of
response associated with the postulated failure mechanisms. Care must be taken
so as not to neglect the critical failure mechanisms that may precipitate collapse
or hazardous debris. Since it is important that the selected analytical approach
be capable of representing the structure’s likely failure mechanism, the designer
must be able to anticipate the response when using simplified analytical methods.
This is not the case for a finite element approach in which the model should
capture all the failure modes.
9.3.1 P-I Diagrams
Pressure-impulse (P-I) diagrams summarize the performance characteristics of
a type of structural or nonstructural element in response to a range of explosive
loading. The explosive loading parameters are expressed in terms of peak pressure and impulse, and the corresponding P-I curves indicate the threshold of the
ANALYTICAL APPROACHES
Figure 9.1
253
Pressure-Impulse (P-I) Diagram for Different Levels of Damage
different levels of protection or hazard. See Figure 9.1. Typically, one end of the
concave P-I curves transitions to a pressure asymptote, while the other end of
the curve transitions to an impulse asymptote. These two asymptotes represent
the minimum impulse for all greater peak pressures or minimum peak pressure
for all greater impulses that correspond to a specified performance threshold. All
combinations of peak pressure and impulse that lie below or to the left of a P-I
curve correspond to a better performance condition, while all combinations of
peak pressure and impulse that lie above or to the right of a P-I curve correspond
to a worse performance condition. The P-I curves either may be derived from an
analytical database of performance or may be fit from available test data. Often,
the parameters are presented in nondimensional form where the peak pressure
and impulse axes are normalized so that a given set of curves and plots may be
used to represent the performance of a greater range of element responses to a
wide variety of explosive threat conditions.
9.3.2 Single-Element Analyses
Individual elements may be analyzed independently by means of either singledegree-of-freedom (SDOF) or multi-degree-of-freedom (MDOF) inelastic dynamic methods. These analytical methods require the use of explicit time-step
numerical integration algorithms, ranging from constant velocity to linear acceleration techniques, along with the appropriate constitutive relations that represent both the elastic and inelastic behavior of the element. SDOF models are
254
STRUCTURAL SYSTEMS DESIGN
inherently simple, and the accuracy of the response calculation depends on the
approximations that are used to characterize the dynamic response of the element. MDOF models may be significantly more complex and may be developed
using finite element methods (FEM) to derive the mass and stiffness matrices.
While it is permissible to consider damping effects, these effects are generally
small in comparison to the large inelastic response that may result from blast
loading and are often neglected.
SDOF SDOF methods are adequately described in a variety of structural
dynamics textbooks (Biggs 1964) and U.S. Army Technical Manuals (Unified
Facilities Criteria Progam 2008b). This approach is based on solving the equation
of motion for a spring-mass system with a nonlinear resistance function using numerical integration. The approximate design methods are based on a RayleighRitz approach for which representative shape functions for the flexural mode are
assumed and used to calculate the effective mass, effective stiffness, and effective
loading function. When the loading pulse is very short relative to the period of the
dynamic system, the initial velocity may be approximated as impulse divided by
the effective mass, and the kinetic energy of the system can be equated to the area
under the force-displacement resistance function. This provides a closed formed
relationship between the required ultimate resistance (Rm) , the applied impulse
(I), the frequency of the dynamic system (ω) and the allowable ductility (µ).
Rm = I ω/(2µ − 1)0.5
The accuracy of these approximate approaches depends on selecting the appropriate SDOF system to represent the governing failure mechanism of the element. While this approach is most commonly used for the design and analysis of
elements that are subjected to blast loading, it requires considerable experience
to make sure a critical failure mechanism is not overlooked. Typically, flexural
modes of response are represented by the SDOF models; however, direct shear
and instability may prove to be more critical response characteristics for shortstandoff detonations and for axially loaded columns. The dynamic loads that may
be applied to SDOF analyses of individual structural elements may be based on
tributary areas or may be derived from the dynamic reactions of the secondary
elements that frame into a primary element. The application of dynamic reactions from subsidiary elements accounts for the frequency of response and limiting capacity of the subsidiary elements, but must also include the appropriate
amount of mass associated with the subsidiary element response. The dynamic
reaction forces from a series of subsidiary element analyses may be converted
into an equivalent dynamic load by equating the work performed by the individual reactions through a representative deformation to the work performed by the
equivalent SDOF response motion. Therefore, if SDOF analyses are performed
for individual secondary beams or purlins, the calculated reaction forces at every time-step may be multiplied by the shape function amplitude of the primary
beam or girder at the location of their connections. If this quantity of work is
ANALYTICAL APPROACHES
255
divided by the integral of the shape function over the span, the resulting quantity
is the effective uniform load at the corresponding time-step that is applied over
the entire span.
MDOF MDOF analyses may be performed using a variety of dynamic response analysis software programs; however, all analyses must be performed
at a sufficiently short time-step to guarantee numerical stability, to resolve the
frequency content of the model, and to accurately represent all potential modes
of failure. Nonlinear dynamic finite element analyses provide the most comprehensive means of determining the self-consistent stiffness and inertial properties,
the limiting plastic limits, and the governing failure mechanisms. Furthermore,
nonlinear geometric algorithms provide the means to include the destabilizing
P-Delta effects due to the coupling of the large lateral transient deformations
and the static gravity loads. MDOF models accurately represent the spatial distribution of dynamic pressures and dynamic reactions from subsidiary element
analyses without having to generate an equivalent dynamic load.
9.3.3 Structural Systems Response
SDOF responses and MDOF models of individual elements do not account for
the interaction between interconnected elements, the phasing of their responses,
or the flexibility of the actual boundary conditions. These simplified approaches
rely on engineering judgment for the selection of approximate boundary conditions, such as soft springs and additional mass. As a result, the idealizations
represented by SDOF and MDOF analyses of individual structural elements do
not accurately account for the dissipation of energy as the entire structure is
deformed and suffers damage. Structural system response calculations include
the relative flexibility and strength of the interconnected structural elements and
provide a more accurate distribution of blast loading. Most importantly, however, structural system analyses consider the phasing of the responses between
the different structural elements. Often, the region of greatest interest is modeled in greater detail, while the rest of the structure is more coarsely represented.
This allows the analyst to dedicate the greatest computational resources to the
response of critical details. However, as with all dynamic response analyses, the
transition from coarse representations to more detailed models must be gradual
so as not to generate spurious results.
9.3.4 Explicit Dynamic Finite Element Analyses
Explicit dynamic finite element methods require sufficient modeling resolution
and short enough time-step to capture the high-frequency characteristics of the
shock wave loading and structural response. Explicit formulations of the equations of motion express the displacement at a given time-step tj + 1 (where the
indices j represent the increments or steps in time for the duration of the response
analysis) in terms of displacements, velocities, or accelerations at previous
256
STRUCTURAL SYSTEMS DESIGN
time-steps. This approach captures the high-frequency effects of the shock loading and easily allows for both material and geometric nonlinear effects; however,
all explicit formulations have a critical time-step, above which the solution becomes numerically unstable. The critical time-step is related to the wave speed of
the material and the least dimension of the finite element model. The equation of
motion is therefore solved at each node, using the current geometry and material
properties at that location and point in time.
An implicit formulation of the equations of motion expresses displacement
of each node at a given time-step tj + 1 in terms of all the displacements, velocities, or accelerations at that time-step tj + 1 . The equations of motion for the
entire system must be solved simultaneously; this is usually done using matrix
methods and is a very efficient way of calculating the response of linear elastic
systems. For nonlinear dynamic systems, however, an explicit formula (predictor) is usually used to estimate the response at the end of each time-step, and this
is followed by one or more corrections to improve the results. Such an approach
is sometimes very inefficient. Furthermore, even for elastic analysis, if the shock
or high-frequency response of the system is needed, the time-step must be made
sufficiently small to capture such effects while avoiding undue numerical damping. Implicit schemes are least efficient in such cases.
Given the complexity of using explicit dynamic finite element methods, great
care needs to be exercised to achieve sufficient accuracy. Verification and validation of both the software and the analyst are critical to the successful completion of high-fidelity first-principle calculations. Comparison of analysis methods
versus physical test data is one method for validating computational tools and
approaches. Benchmarking versus known analytical solutions or against other
software packages provides further confidence in software and methods. Unless
established methods are used based on validation exercises, finite element methods need to be checked with grid refinement calculations to determine whether
the model has sufficient resolution. Fine resolution and detailed models are no
guarantee of accurate results. Accurate representation of material response, sufficient physical extent of the computational model, and proper boundary conditions are critical to calculating accurate results.
9.4 PROGRESSIVE COLLAPSE
The protective design of structures is often referred to as threat-dependent or
threat-independent. In one case, a structure is designed to provide a specified
level of performance in response to a specific design basis threat, while in the
other, the structure is designed to be insensitive to local damage resulting from
unforeseen extraordinary events. In many cases, the uncertainty of the maximum credible threat or the inability to accurately characterize the performance
in response to explosive loading requires both approaches to be applied to the
design. While the threat-dependent approach is related to the design basis threats,
the threat-independent methods evaluate the potential for an initiating event to
PROGRESSIVE COLLAPSE
257
precipitate a broader extent of collapse. The prevention of progressive collapse
may therefore be achieved through the hardening of key structural elements in
order to preclude the initiating event, through the provision of ductile redundant
load paths, or through compartmentalization that permits local failure to occur
without endangering adjoining structural elements. In all cases, consideration
must be given to both global ductility demands that involve all elements of the
framing system and local ductility in the vicinity of the initiating damage.
The insensitivity to local failure, often referred to as robustness, is a property
of the structure, and is independent of possible causes of initial local failure.
Where specific threats are not prescribed, or for structures that are deemed low
risk, an assumed extent of initial local failure and assumed extent of damage
to the remaining structure may be prescribed. For these structures, element-toelement connectivity, minimum tie forces, the elimination of nonductile failure
mechanisms, and enhanced stability requirements may improve the structure’s
robustness. If the structural members and connections can develop these minimum tie forces, which vary with construction type and location in the structure,
the structure will be held together in the event local damage disrupts the primary
load paths. The purpose of the horizontal and vertical ties is to enhance continuity and ductility, and to develop alternate load paths in the structure. Internal
and peripheral horizontal ties are typically provided, along with ties to external
columns and walls. The extent of tie force requirements may either be prescribed
or calculated, starting with an assumed damage pattern. The alternate load path
requirements, also termed “bridging requirements,” may be determined using dynamic nonlinear analytical methods and known material damage limits, or may
be approximated using equivalent static elastic analysis methods with nominal
material damage criteria. The level of conservatism may be varied depending on
the level of approximation or the risk assessment for the structure.
Where the characteristics of the structure make bridging impractical, local
hardening of key elements or the introduction of structural fuses may be introduced into the structure to either prevent the initial damage or define a limited
zone of collapse. For these types of structures, the level of specific local resistance for key elements or the zones of limited collapse must be defined.
Where a threat and risk assessment quantifies the hazards to which a structure may be subjected, specific initiating events may be prescribed, and the performance of the structure may be calculated. Nonstructural protective measures
may be developed to deny access to key structural members, or details and sacrificial elements may be introduced to isolate the critical components. For this
performance-based approach, the structural system redundancy and element capacities would be designed to resist these initiating events or resist the progression of damage or collapse. The increased resistance is therefore based on a prescribed set of actions and a permissible extent of resulting damage. The intensity
of the initiating actions and the permissible response limits may be specified in
the security performance criteria for the structure. The analytical methods that
demonstrate the behavior of the structure must accurately represent the dynamic
nature of the event and the damage-state behavior of the structural materials.
258
STRUCTURAL SYSTEMS DESIGN
Regardless of whether local hardening, bridging over a damaged structure, or
compartmentalization is used, a post-event stability analysis of the structural system must be performed.
For blast-resistant design, a threat must be identified in order for the initial
state of damage and the residual capacity of the structure to be determined. This
approach is similar to that of performance-based seismic design codes for which
the engineer selects the hazard (event) and the desired performance, such as low
levels of damage (immediate occupancy), in response to a small earthquake and
life-safety protection (damage short of complete collapse) in response to an extreme event. Either the structural elements must be designed to resist the specified threat and maintain load-carrying capacity, or a finite extent of localized
damage may be permitted, as long as the damaged structure is still capable of
preventing a progression of collapse. This corresponds to the “extreme event” in
response to which the structure must remain standing for life safety and evacuation of the occupants.
9.4.1 European Guidance
In the Euro-Norm EN 1991-1-7 (British Standards Institute 2006), the strategies
for preventing collapse consider two different approaches: strategies based on
identified accidental actions and strategies based on limiting the extent of localized failure. The strategies based on accidental actions consider designing the
structure to have sufficient minimum robustness, protective measures that prevent the action or reduce its effects, or designing the structure to resist the effects
of the action. The strategies based on limiting the extent of localized failure address the “threat-independent” prescriptive approaches for robustness, such as
enhanced redundancy through alternate load paths, key element designs to withstand notional accidental action, and prescriptive rules for integrity and ductility.
However, the criteria for protection are specific to the project and are determined through consultation with the client and the relevant authority. Since
strategies based on unidentified accidental actions cover a wide range of possible events, the means to prevent progression of collapse are typically based on
limiting the extent of localized failure. The magnitudes of accidental actions that
are identified, as in the case of internal explosions and impact, are specified in
the EuroNorms. In some cases where there is no risk to human life and where
economic, social, or environmental consequences are negligible, the complete
collapse of the structure caused by an extreme event may be acceptable.
9.4.2 U.S. Guidance
All current U.S. Guidance documents are based on the U.K. standards that were
based on the partial collapse of the Ronan Point apartment house. These standards extrapolated from the behavior of a panelized precast structure to develop
prescriptive and performance-based guidelines for other forms of construction.
Similar requirements form the basis for the most recent Building Code of the
PROGRESSIVE COLLAPSE
259
City of New York (New York City Department of Buildings 2008), the 2008 Supplement to the International Building Code (International Code Council 2008),
and the U.S. Department of Defense’s UFC 4-023-03 (Unified Facilities Criteria 2008a). Since these approaches all rely on catenary action, the structural
detailing must consider connections that are able to undergo significant inelastic
rotations while sustaining large tensile axial loads.
Since the alternate path method is inherently dynamic and inelastic, the
optimal analytical approach considers material and geometric nonlinearity and
inertial effects. While the analysis of dynamic systems may be performed using
implicit or explicit numerical integration algorithms, the introduction of material
and geometric nonlinearity complicates the analysis. Implicit schemes require
the reformulation of the stiffness matrix whenever properties are significantly
altered, and this can become computationally inefficient when significant extents
of inelasticity or catenary behavior are introduced. Explicit schemes require
small enough time-steps to ensure numerical stability, and this too may become
computationally inefficient. Furthermore, unless the beam-column joints and
connections are explicitly modeled, their behavior must be approximated with
equivalent nonlinear rotational and translational springs. The force displacement
relations for these equivalent nonlinear springs are best determined by performing comprehensive component analyses of specific joints, as was done for the
analysis of the Deutsche Bank Building for the World Trade Center Building
Performance Study (Federal Emergency Management Agency 2002).
The most recent version of UFC 4-023-03 (UFC, 2008a) proposes a simplified
procedure that is based on the approaches established for Seismic Rehabilitation
of Existing Buildings (American Society of Civil Engineers 2007). A ductile
material that develops significant deformation following yield, over which the
force is either constant or increasing, is said to exhibit deformation-controlled
behavior. Alternatively, a brittle material that experiences a steep reduction
in capacity following limited inelastic deformation exhibits force-controlled
behavior. The representations of structural behavior that are reported in ASCE
41-06 are considered to be reasonable for the evaluation of damage-state performance of buildings. Although simplifications to the dynamic inelastic response
analyses reduce the extent of computational resources needed to calculate the
performance of the structure following the removal of a key element, these
simplified methods require detailed guidance in order to capture all reasonable
failure mechanisms. UFC 4-023-03 offers two simplified static analysis procedures, linear static procedure (LSP) and nonlinear static procedure (NSP), as
alternatives to the nonlinear dynamic procedure (NDP) approach. These simplified approaches allow the use of commercially available structural analysis
software that is readily available in structural design offices. The simplifying approximations allow the analyst to apply load factors to static analyses in order to
account for the missing dynamic effects, and provide the means to interpret the
results of linear analyses to account for the material and geometric nonlinearities. Since the NDP is the most accurate representation of the structural response
to an initial state of damage, there is little ambiguity to the interpretation of the
260
STRUCTURAL SYSTEMS DESIGN
calculated results. Correspondingly, all approximate methods are necessarily
conservative, and a more representative analytical understanding will allow the
designer to achieve a specified level of robustness using a wider variety of structural options. The use of the simplified LSP is restricted to buildings with regular
framing plans, which are defined as irregularity limitations. Both approximate
methods were calibrated to nonlinear dynamic analysis to establish simplified
system-dependent factors that account for the anticipated dynamic response.
The LSP may be used wherever structural irregularities and demand capacity ratios conform to prescribed limitations. For the LSP, the structural model
is analyzed in response to both deformation and force-controlled actions. The
actions applied to the tributary areas above the removed key element are amplified to account for the inertial effects associated with the sudden removal of
the column, and a concurrent lateral load is applied individually to each face
of the structure to evaluate stability. The LSP may still be used if the structural
framing is irregular, provided the calculated demand/capacity ratios (DCR) are
less than or equal to 2.0; otherwise the NSP must be used. Once the acceptable use of the LSP is determined, the potential for disproportionate collapse
is calculated using component or element demand modifiers (m-factors) in an
approach that is similar to that of ASCE 41-06. M-factors are tabulated for different types of structural systems and details for different levels of damage-state
performance. These m-factors are used to determine both the Dynamic Increase
Factors and the maximum acceptable forces for deformation-controlled actions.
The expected material strength and the specified material strength respectively
define material properties for deformation-controlled and force-controlled analyses, and the expected component strength (QCE ) and lower bound estimate of the
component strength (QCL ) respectively define action capacities for deformationcontrolled and force-controlled analyses. QUD is defined as the deformationcontrolled action, from linear static model; QUF is the force-controlled action,
from linear static model; and is the strength-reduction factor from the appropriate material-specific code.
m Q CE ≥ Q UD deflection-controlled criteria
Q CL ≥ Q UF force-controlled criteria
The use of the NSP is not restricted by either irregularity or DCR limitations. The actions applied to the structure for the NSP are the same for
calculating displacement-controlled and force-controlled responses; however,
the force-displacement behavior of all structural components must be explicitly
modeled, including connections wherever they are weaker or less ductile than
the connected components or the flexibility of the connection produces more
than a 10% change in the connection forces or deformations. Tabulated plastic
rotation angles, as in ASCE 41-06, are used in the NSP to determine the
Dynamic Increase Factors; the deformation limits determine compliance for
deformation-controlled actions, and the lower bound estimate of the component
strength determines compliance for force-controlled actions.
REFERENCES
261
While the UFC relies on the best information available, based on seismic
research, the justification for the specified factors and procedures requires rigorous experimental and analytical study. Furthermore, the performance of actual structural details for the different structural materials and systems requires
extensive analytical research and development. Although simplified procedures
must conservatively represent inelastic material properties and large displacement mechanics, there is little justification for overly precise procedures to approximate the effects of dynamics and nonlinear behavior. Furthermore, the correlation between the simplified analytical representation of structural behavior
and the actual performance of structural components undergoing large deformations requires significant testing and analysis. Even the use of conventional
understrength factors, factors, may be questioned for the regime in which materials will be severely distorted.
REFERENCES
American Society of Civil Engineers. 2007. Seismic Rehabilitation of Existing Buildings
(ASCE 41-06). New York: American Society of Civil Engineers.
ASTM. 2003. Standard Practice for Specifying an Equivalent 3-Second Duration Design
Loading for Blast Resistant Glazing Fabricated with Laminated Glass (ASTM F2248).
West Conshohocken, PA: ASTM International.
ASTM. 2007. Standard Practice for Determining Load Resistance of Glass in Buildings
(ASTM E1300). West Conshohocken, PA: ASTM International.
Biggs J. M. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill Book
Company.
British Standards Institute. 2006. Eurocode 1: Actions on Structures—Part 1–7: General
Actions—Accidental Actions (EN 1991-1-7:2006). London: British Standards Institute.
Federal Emergency Management Agency. 2002. World Trade Center Building Performance Study: Data Collection, Preliminary Observations and Recommendations
(FEMA 403). Washington, DC: Federal Emergency Management Agency, Department
of Homeland Security.
Hyde D. W. 1992. CONWEP, Conventional Weapons Effects (USAEWES/SS-R).
Vicksburg, MS: U.S. Army Engineer Waterways Experiment Station.
International Code Council. 2008. Supplement to the International Building Code.
Washington, DC: International Code Council.
Kingery C. N. and G. Bulmash. 1984. Airblast Parameters from TNT Spherical Air
Burst and Hemispherical Surface Burst (ARBRL-TR-02555, AD B082713). Aberdeen
Proving Ground, MD: U.S. Armament Research and Development Center, Ballistic
Research Laboratory.
New York City Department of Buildings. 2008. Supplement to the Building Code of
the City of New York, Local Law No 76 Effective December 6, 1968. New York:
Department of Buildings.
Oswald C. J. 2005. Component Explosive Damage Assessment Workbook (CEDAW)
Methodology Manual V1.0 (02-0752-001), San Antonio, TX: Baker Engineering and
Risk Consultants, Inc.
262
STRUCTURAL SYSTEMS DESIGN
Protective Design Center. 2006. Methodology Manual for the Single-Degree-of-Freedom
Blast Effects Design Spreadsheets (SBEDS) (PDC-TR-06-01), Omaha, NE: U.S. Army
Corps of Engineers, Protective Design Center.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (UFC 3-340-01). Washington, DC: Department of
Defense, Unified Facilities Criteria Program.
. 2008a. Design of Buildings to Resist Progressive Collapse (UFC 4-023-03).
Washington, DC: Department of Defense, Unified Facilities Criteria Program.
. 2008b. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: Department of Defense, Unified Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
10
Building Envelope and Glazing
Eve Hinman and Christopher Arnold
For an explosion outside a building, the exterior envelope is the critical line of defense that separates the people, operations, and contents inside the building from
the air-blast effects outside the building. Unfortunately, for most standard building types, the building envelope is ill-suited to resisting air-blast loads, which are
enormous and which act directly on the surface area of the exterior envelope in
the out-of-plane direction.
In this chapter, some of the key concepts needed to optimize the design of
the exterior envelope to mitigate explosion effects are presented. A variety of
wall and window types are considered, as well as doors and louvers. Roof and
foundation systems are also addressed.
10.1 DESIGN INTENT
The primary design objective for the exterior envelope of most standard building
types, is to mitigate the hazard of flying debris generated by failed exterior walls,
windows, and other components, to reduce casualties and business disruption and
facilitate rescue and evacuation efforts.
Other objectives are to design the exterior envelope to:
r Fail in a way that does not initiate progressive collapse
r Keep the air blast outside the building to the best of its ability
10.1.1 Life Safety
Impact of flying debris entering a building causes laceration and blunt trauma
injuries, and potentially fatalities. Once the exterior envelope is pierced and the
air blast enters the building, injuries such as eardrum rupture, lung collapse, and
blunt trauma and concussion due to being thrown against objects are the dominant injury modes. Impacts due to internal nonstructural damage to partitions
and ceiling, light fixture, and equipment components and falling furniture are
also a hazard.
263
264
BUILDING ENVELOPE AND GLAZING
Although it is cost-prohibitive to eliminate all these hazards, it is possible to
reduce the risk to a level that is acceptable through informed design decisions.
10.1.2 Emergency Egress and Facilitating Search and Rescue
The building exterior envelope design may be optimized to make it easier for
people to get out and for emergency responders to enter safely.
After an explosive event, the debris created by the failed exterior envelope
can exacerbate casualties by reducing the efficiency of evacuation and rescue
operations. Some examples of this are:
r
r
r
r
Internal debris can cause occupants to become trapped in rubble piles.
Exterior debris can block emergency exits.
Debris along egress paths can cause delays and additional injuries.
Exterior falling fragments post-event trigger safety precautions for emergency operations, which can significantly delay rescue, thereby putting the
injured at greater risk.
r Toxic dust or fumes generated by debris may have long-term health ramifications.
10.1.3 Critical Functions (Protecting Equipment and Business Processes)
Even at low pressures, air blast that enters the building can cause extensive nonstructural damage to interior partitions and suspended ceilings. The resulting
internal disarray can cause extensive business disruption. Even if the damages
are repairable, total cost required to return the building to pre-event functionality
may exceed the replacement cost of the building.
Sensitive mechanical parts of critical equipment can be disabled by air blast
or flying debris that enters the building through the exterior envelope. An interior
baffle wall may be used to dissipate the energy of the air blast by forcing it to
change direction two or more times before it reaches the equipment. Some other
options are to use storm louvers or forced-entry-resistant louvers, which have
a chevron cross section which are able to perform a similar function. Louver
attachments into the structure are to be designed to resist the capacity of the
louver to ensure that the louver fails before the connections. In some cases, a
catch system using a grate securely fastened behind the louver may be used to
prevent impact into delicate equipment components.
Critical equipment placed on the roof will be less vulnerable to a ground level
explosion than equipment at or below street level. For roof equipment, a hardened wall with roof may be used for protection. However, for this solution, note
that if the wall fails, the impact of debris may have a more detrimental effect than
using lightweight housing that is not blast-resistant.
It is recommended that the transformer be placed away from vulnerable areas,
for instance in a parking lot, near the loading dock, or under the sidewalk. Inside
DESIGN APPROACH
265
the building in a protected location is preferred. The switchgear room is to be
separated to the extent practical from the transformer(s), the generators, and any
other backup systems, such as an uninterruptible power supply (UPS) or batteries, to improve the chances of retaining at least emergency functions in the event
of an explosive attack. In addition to separation, redundancy of critical systems
is highly recommended for higher-risk buildings. An example of this is placing
one set of generators on one side of the building, and another set on the other
side of the building to create redundancy.
10.2 DESIGN APPROACH
For buildings with a low to moderate risk of explosive attack, the design
approach is to first design the building exterior envelope for conventional loads,
and then to evaluate the response to explosive loads. If the design team is
properly consulted when key decisions are made during the concept phase, then
the modifications required to meet explosive load requirements can be addressed
efficiently. This approach ensures that the design meets all the requirements for
gravity and natural hazards, in addition to air-blast effects, and is effective for
buildings where explosion effects are a secondary concern.
Take note that some features that are beneficial for blast resistance diminish
performance for other loads, if not properly integrated. For example, using a
plastic coating that emits toxic fumes in the event of a fire may not be a good
solution for enhancing the blast performance of an exterior wall, since fire is a
more common hazard than terrorist attack.
Another example that presents a potential conflict between the seismic design
and blast design is that increased mass generally increases the design forces,
whereas for explosion loads, mass generally improves response. This is because
of the highly impulsive nature of explosive loads. Heavy structural components
have longer periods of vibration so that, by the time the mass, M, is mobilized and
the structural component begins to move with a velocity, V, the load or impulse,
I, has been removed. In mathematical terms, the kinetic energy, KE = MV 2 /2, or
momentum, MV, imparted to the structural component is directly proportional
to impulse and inversely proportional to the mass: i.e., KE = MV 2 /2 = (Mv)2 /
(2M) = I 2 /(2m). In this situation, careful coordination between the blast consultant and the structural engineer is needed to provide an optimized response that
accounts for both loading cases.
Good antiterrorism design is a multidisciplinary effort, requiring the concerted efforts of the architect, structural engineer, security professional, and the
other design team members. For instance the passive protection provided by
effective building design needs to be balanced with the active or operational
security measures implemented (e.g., guards, cameras, dogs, sensors, magnetometers). It is critical that these two sets of measures complement and support
one another. In particular, the size weapon that is considered at the perimeter
and at the exterior envelope is a function of the size weapon that is able to be
266
BUILDING ENVELOPE AND GLAZING
easily detected through operational security measures. Like protective design,
security needs to be implemented early in the design to be effective. Lines
of sight; lighting and placement of cameras along the exterior envelope; and
location, number and type of entrances provided are examples of security issues
to be addressed early in design with the architect, since they potentially affect
the design threat to be used.
The envelope system is to be designed by keeping in mind the concepts of
balanced or capacity design (see Section 10.2.3) and ductile response. Ductile
systems are able to absorb energy through inelastic flexural deformation. Systems that fail in a brittle manner through shear are to be avoided. By permitting
limited exterior envelope damage that does not significantly increase the hazard to the majority of the occupants, it is possible to design more cost-effective
systems that absorb energy through deformation, and transmit lower forces into
the connections and supporting structure, thus reducing the potential for more
serious structural failures.
Perhaps the simplest analytical method is to use a pressure-impulse, or P-I,
diagram, in which various damage levels for a component with a given resistance are expressed as curves, and the axes are peak pressure and impulse. By
plotting the pressure and impulse acting on the system, you can estimate the
level of damage to the exterior envelope component and compare it with allowable limits. These curves may be based on explosive tests or on the output from
a semi-empirical or a single-degree-of-freedom system. The limitation of these
tools is that they are based on constant parameters for the component under consideration. If the component that is being investigated differs in any way, some
extrapolation or interpolation between charts is needed.
Typically, simplified or single-degree-of-freedom (SDOF) methods used for
the design of structural components may be used for the exterior envelope (Unified Facilities Criteria Program 2008). For these systems, the component timedependent peak deflection, X(t), is evaluated by modeling the component as a
lumped mass, M, so that the inertia force is MA(t) where A(t) is the acceleration
or second derivative of X(t). The resistance is provided by a spring, R(x), which
has a linear elastic flexural stiffness, K, and a ultimate flexural resistance of Ru .
The loading is modeled as a time-dependent function, F(t). For a system initially
at rest, the resulting equation of motion is:
M A(t) + R(x) = P(t)
where: R(x) = KX(t) when X(t) is less than or equal to the maximum elastic limit
displacement Xe and
R(x) = Ru when X (t) > X e
In these expressions, M and R, and P are “equivalent” values which represent the portion of the mass, resistance, and loading that is participating in
the displacement of the member, and which provide a natural frequency and
DESIGN APPROACH
F(t)
267
R
RU
KLMM
K
1
Xe
R(x)
X
Xm
F
Fl
td
f
Figure 10.1 Graphical Representation of an SDOF System
displacement that are the same as that of the actual system. Figure 10.1 provides
a graphical representation of an SDOF system.
These equations may be easily solved by using numerical methods. The
resulting maximum displacement is typically evaluated by comparing the ratio
of the maximum displacement Xm and the maximum elastic limit displacement,
or ductility Xm/Xe. Another measure is the support rotation, which is defined
as the angle between the undeflected member and the line connecting the
support point to the point of peak deflection. For example, for a uniformly
loaded, simply supported beam of length L the support rotation is equal to the
tangent of 2Xm/L. Figure 10.2 illustrates the concepts of displacement ductility
and rotation. Typically, the maximum allowable ductility may vary from 1
for a brittle material like monolithic glass to about 10 for ductile systems like
steel. For heavily damaged systems, a ductility of up to 30 may be allowed.
Permitted ductilities are generally about 2 degrees for most systems, but may be
permitted to be 4 degrees or more for highly ductile structures where extensive
but low-hazard damage is permitted.
Other models that may be used include two or more degrees-of-freedom systems or piecewise linear continuum models. Two-degrees-of-freedom models
have been used with some success, in particular for a beam and girder where
a portion of the load is transmitted directly to the girder from the beam, and
where the beam experiences rigid body as well as deformational displacements.
Piecewise linear systems are of particular interest because they are able to represent behavior more accurately for bilinear or trilinear resistance functions and
loading functions. This type of analysis was not a practical solution until the
recent advent of powerful computational tools such as MATLAB (MathWorks
2006) or Mathcad (MathSoft 2001).
However, for complex exterior envelope systems or air-blast loading scenarios
or systems using innovative materials or energy absorption methods, simplified
268
BUILDING ENVELOPE AND GLAZING
R
RU
K
X
Xm
Xe
L
Xm = Peak Displacement
Plastic Hinge
Figure 10.2 Graphical Representation of the Concepts of Displacement Ductility and
Rotation
analysis is not sufficient. In these cases, the engineer needs to use finite-element
numerical time integration techniques and/or explosive testing to verify response.
Again, nonlinear dynamic time history analysis is required to adequately model
the system. Material models used for these programs are generally more sophisticated than the linear elastic perfectly plastic models used for simplified analysis. In this case, the material models provide a more realistic representation
for large deformation. Examples of computer codes on the market that are capable of performing these types of computations include SAP2000/NonLinear
(Computers and Structures 1990) and LS-DYNA (Livermore Software Technology Corp. 1999). There are also codes available to model the expansion of an air
blast and its reflection against the structure surface, which may be used to generate the loads applied to the model using a software product such as CONWEP
(Hyde 1988) or ATBLAST, a software product prepared by Applied Research
Associates.
The time and cost required to perform the analysis cannot be ignored in choosing computational methods. Because the design process may require iteration,
the cost of analysis must be justified in terms of benefits to the project and increased confidence in the reliability of the results. In some cases, a simplified
approach will be used for the preliminary design and a more sophisticated approach, using finite elements and/or explosive testing, may be used for the final
verification of the design.
If testing has been performed for a similar system, it is necessary to obtain the
testing report and verify that the assumptions and findings are consistent with
the project requirements, prior to accepting testing in lieu of detailed analysis.
In particular, check to make sure that the support conditions for the member are
DESIGN APPROACH
269
accurately represented. Also check to see that both the pressure and impulse are
at least as large as what the design requires. If only the pressure or the impulse
meets criteria, some additional calculations will be needed to verify response.
It is recommended that the testing be supplemented with pre- and post-analysis,
using an advanced method such as finite elements.
Often, only a proof of concept design is provided by the blast engineer during
design. The final blast design is the responsibility of the subcontractor implementing the design. A detailed set of specifications is strongly recommended
to ensure that the design intent is met. Some of the requirements to consider,
including in the specifications for the exterior facade subcontractors, are:
r Provide performance-based design requirements for response.
r Put all the blast requirements in a single location in the specifications, rather
than sprinkling them through the individual sections. This makes it easier
for everyone to keep track of these special requirements.
r Require the blast engineer’s approval of the subcontractor based on successful performance on similar projects.
r Require the subcontractor to submit blast calculations for review.
10.2.1 Response Criteria
Levels of damage to the exterior envelope may be described by these terms:
minor, moderate, major, catastrophic. Qualitative descriptions of each damage
level are given below as they pertain to the exterior envelope. Note that all these
descriptions imply that there is little or no structural damage (i.e., all the damage
is limited to the exterior envelope).
Minor: There is little if any damage. The exterior envelope may sustain some
damage, but virtually no failure. Damage may consist of cracked windows,
failed sunshades still attached to the building, or permanent deformation
of walls and window frames. Injuries may be expected, and fatalities are
possible but unlikely.
Moderate: Failure of the exterior envelope is confined to a localized area and
is usually repairable. Hazardous window failure is limited to one face of
the building. Injuries and possible fatalities are expected.
Major: Extensive exterior envelope damage and some roof damage are expected. In this case, extensive fatalities are expected. Damage is usually
not repairable.
Catastrophic: In this case, all or the majority of the exterior envelope has been
removed.
Generally, moderate damage for the design basis threat is a reasonable goal
for new construction to meet life-safety objectives. For buildings that need to
270
BUILDING ENVELOPE AND GLAZING
remain operational during and after an event, or are designated as high risk,
minor damage may be the more appropriate damage level.
Quantitative performance is usually evaluated by comparing the ductility (i.e.,
the peak displacement divided by the elastic limit displacement) and/or support
rotation (the angle between the support and the point of peak deflection) to empirically established maximum values that have been established by the military
through explosive testing. Maximum permissible values vary depending on the
material and the acceptable damage level. Some criteria documents do provide
the design values that need to be met.
10.2.2 Static versus Dynamic
A dynamic nonlinear approach is more likely to provide a design that meets the
design constraints of the project than a static approach. Elastic static calculations are likely to give overly conservative design solutions if the peak pressure
is considered without the effect of load duration. By using dynamic calculations
instead of static, we are able to account for the very short duration of the loading. In addition, the inertial effect included in dynamic computations greatly improves response. This is because by the time the mass is mobilized, the loading
is greatly diminished, enhancing response. Furthermore, by accepting that damage occurs, we are able to account for the energy absorption of ductile systems
that occurs through plastic deformation. Finally, because the loading is so rapid,
we are able to engage the enhanced material strength that often occurs with very
high strain rates.
10.2.3 Balanced Design
Balanced or capacity design philosophy for a building system refers to designs
that are controlled such that the failure of the weakest component in the system
results in the least destruction. For a window system, for instance, it refers to the
window glass failing at pressure levels that do not exceed those of the frame, anchorage, and supporting wall system. If the glass is stronger than the supporting
members, then the window is likely to fail with the whole panel entering into
the building as a single unit, possibly with the frame, anchorage, and the wall
attached. This failure mode is considered more hazardous than one in which
the glass fragments enter the building, provided that the fragments are designed
to minimize injuries. By means of a damage-limiting approach, the damage sequence and extent of damage are controlled.
10.2.4 Load Path
Load path in the exterior envelope generally refers to the transmission of load
from the member that presents the most surface area to the exterior (i.e., glass
pane, panel, wall, slab) into the building structure. Connections and reaction
forces, two of the more nontrivial parts of the path, are discussed below.
DESIGN APPROACH
Connections
271
For the exterior facade, the connections include:
r The connection between the window glass and the window mullion, transom or frame
r The connection of the window mullion or transom and the window frame
r The connection between the window frame and the wall
r The connection between the wall and the structural frame
Connections are typically the least ductile elements in the exterior envelope
system and the most prone to failure if not properly designed. The connections
are required to enable the supporting elements to reach their ultimate flexural
capacity without failure. They also need to remain intact for the loads imposed
by large inelastic deformations of the supporting members, which in turn can
cause significant deformation of the connections.
Fasteners holding the window frame in the wall tend to be placed into tension
and shear as the window undergoes deformation. The wet glazing are put into
shear as a laminated glass pane begins to act as a membrane. As the mullion of a
unitized system deforms, stresses are placed on the connections holding the two
halves of the mullion together.
The failure envelope of connections respondingto load combinations such as
tension and shear need to be considered as part of the design.
Although it seems as if window frame connections should be routinely designed to carry the ultimate loads transmitted by the window pane, this is not the
case. The window frame is designed to carry the design loads only. The capacity
of the glass may be significantly higher than the design loads.
Reaction Loads (Ru versus Applied Load) Reaction loads may conservatively
be approximated by the static ultimate capacity of the loading element, and by
setting the load and strength factors for the connection design to 1.0. The load
factor is set to 1 because we are assuming that the supported member has been
pushed to its capacity. The reaction load cannot be higher than the capacity of the
member. Similarly, the strength factor is set to 1, again in recognition of the fact
that we are letting the supporting member reach its ultimate capacity. Although
this may not seem conservative, it is, because we neglected the mitigating effects
of ductility and inertia on the system.
This approach is consistent with capacity or balanced design principles and
is the best and most reliable. However, it can lead to designs that are not constructible or affordable. If this is the case there are several approaches which
may be considered:
r Use the average applied load instead of the peak load, because the edge
load along a panel transfers a load that varies with its peak value at the
center point and decreases towards the ends. Often the peak value is given,
but the average load along the length is less and is more representative of
the loading on the anchors.
272
BUILDING ENVELOPE AND GLAZING
r Increase the strength factor for flexural or tensile response, to account for
strain-hardening effects. For impulsively applied loads, the strength of the
material is increased. This effect is referred to as strain hardening.
r Permit the loss of a secondary member (for instance, permit the loss of a
beam but not a girder). By enabling the major element in the load path to
be designed to the capacity of the elements it supports, we are improving
performance while maintaining the cost-effectiveness of the design. This is
done often for roof systems where the floor slab capacity load is greatly in
excess of the design loads.
r Use an element with a lower ultimate capacity. This is a practical and effective tool for systems where there is significant ductility. For instance, a thin
precast panel will have a lower ultimate resistance than a thicker panel and
can be designed to perform well in ductility.
10.3 FENESTRATION
Windows, once the sole responsibility of the architect, become a structural issue
once explosive effects are taken into consideration. Windows designed to mitigate the effects of explosions are first to be designed to resist conventional loads,
and then to be checked for explosive load effects and balanced design.
Even if the building remains standing and no structural damage occurs, extensive injuries can occur due to nonstructural damages. Windows are typically
the most vulnerable portion of any building. Although it may be impractical to
design all the windows to resist a large-scale explosive attack, it is desirable to
limit the amount of hazardous glass breakage to reduce the injuries. Typical glass
windows break at low pressure and impulse levels, and the shards created by broken windows are responsible for many of the injuries incurred due to large-scale
explosive attack.
Designing windows to provide protection against the effects of explosions can
be effective in reducing the glass laceration injuries. Glass lacerations are caused
when the shards of glass created when a window is broken are propelled into the
building and penetrate the skin. This is perhaps the most common type of injury
that results from large-scale explosion incidents and may occur hundreds or even
thousands of feet from the source in urban areas, where the air blast attenuates
more slowly as it propagates through the street “canyons.”
Window protection should be evaluated on a case-by-case basis by a qualified
protective design consultant, to develop a solution that meets established objectives. In the sections below, a number of generic recommendations are given for
the design of the window systems to reduce injuries to building occupants.
To limit glass laceration injuries, there are several approaches that can be
taken. One approach is to reduce the number and size of windows, which
potentially will reduce the air-blast and glass shards entering the building,
FENESTRATION
273
thus reducing the interior damage and injuries. Specific examples of how to
incorporate these ideas into the design of a new building include:
r Limiting the number of windows on the lower floors where the pressures
are higher due to an external explosive threat
r Using an internal atrium design with windows facing inward toward the
atrium, rather than outward toward the facility perimeter
r Windows, which are close to the ceiling, above the heads of the occupants
r Angling the windows away from critical weapon locations to reduce the
pressure levels
10.3.1 Glass
Glass is often the weakest part of a building, breaking at low pressures compared to other components such as the floors, walls, or columns. Past incidents
have shown that glass breakage and associated injuries may extend many thousands of feet in large external explosions. High-velocity glass fragments have
been shown to be a major contributor to injuries in such incidents. For incidents
within downtown city areas, falling glass poses a major hazard to passersby and
prolongs post-incident rescue and cleanup efforts by leaving tons of glass debris
on the street.
As part of the damage-limiting approach, glass failure is not quantified in
terms of whether breakage occurs or not, but rather by the hazard it causes to the
occupants. The glass performance condition is defined based on empirical data
from explosive tests performed in a cubical space with a 10-foot dimension. The
performance condition ranges from 1, which corresponds to no glass breakage,
to 5, which corresponds to hazardous flying debris at a distance of 10 feet from
the window.
Design criteria established by the Interagency Security Council, which governs the antiterrorism efforts for most agencies, require performance condition
3b for buildings that are at moderate risk of attack. At this protection level, the
window breaks, and fragments fly into the building but land within 10 feet of the
window. The design goal for moderate-risk buildings is to achieve a performance
level of less than 4 for 90% of the windows. Protection level 4 means that glass
fragments will penetrate a witness panel located 10 feet behind the window for a
height of two feet or less.
The preferred solution for new construction is to use laminated glass with
structural silicone sealant (i.e., wet glazing) around the inside perimeter. The
lamination holds the shards of glass together in explosive events, reducing its
potential to cause laceration injuries. The structural sealant helps to hold the
pane in the frame for higher loads. For insulated units, only the inner pane needs
to be laminated because the inner pane will provide protection against the hazard
of shards generated by the outer pane.
274
BUILDING ENVELOPE AND GLAZING
Annealed glass has a breaking strength that is about one-half that of heatstrengthened glass and about one-fourth that of tempered glass (Amstock 1997),
thus reducing the loads transmitted to the supporting frame and walls. To reduce
reaction loads, it is sometimes advantageous to consider annealed glass. The
preferred interlayer thickness is 60 mil (0.06 in) unless otherwise specified by
the criteria. At a minimum, 30 mil is to be used.
To make sure that the components supporting the glass are stronger than the
glass itself, we specify a window breakage strength that is high compared to
what is used in conventional design. The breakage strength in window design
may be specified as a probability of the number of windows expected to break
at that load. For instance, when designing for conventional loads, it is typical
to use a breakage pressure corresponding to a probability of 8 breaks out of
1,000. Where there is a high probability of extensive glass breakage, as in an
explosion incident, a pressure corresponding to 750 breaks out of 1,000 is used.
Glass breakage strengths may be obtained from window manufacturers.
Smaller glass panes generally have higher capacities than larger panes. Consequently, smaller panes can cause significantly higher loads to be transmitted
to the frames than larger panes. As a result, in blast-mitigating design, we avoid
small panes. Also, since every size pane has a difference capacity, it is desirable
to standardize the design as much as is practical.
There are several government-sponsored software products available for evaluating the response of window glass for use on federal projects, including HAZL
(U.S. Army Engineer Research and Development Center 2001), WINGARD,
and WINLAC. These codes are made available to government contractors who
have government projects requiring this type of analysis. One approach that
is publicly available is described in the criteria by the U.S. military (Unified
Facilities Criteria Program 2007), using ASTM Standards (ASTM 2003,
2004a). For those who are designing unique systems for specific threats not
covered in criteria documents, it is recommended that the system be tested
(ASTM 2001).
Glass block is generally not recommended because of the heavy projectiles
these walls may create due to failure at the mortar lines. However, blast-rated
glass-block products are available in which the glass blocks are framed by a
steel grate system or other method.
The way in which the glass is supported into the exterior envelope can have
an effect on the severity of the glass hazard. A brief summary of various types of
supports is given below.
Punched Windows: Punched windows (see Figure 10.3) often can be designed
efficiently to be less vulnerable than other window types because the frame is
attached directly into the wall, which is generally of more robust construction.
Strip Windows: Strip or ribbon windows (see Figure 10.4) generally consist
of alternating bands of glazing and opaque material constructed using precast
panels, poured concrete, or insulated metal panels. The opaque area conceals
the floor structure and is referred to as a “spandrel.” Ribbon windows have thin
vertical metal framing separating the individual panes, and are supported by
FENESTRATION
275
Figure 10.3 Example of “Punched” Windows
the wall at the top and bottom. For this type of exterior facade, the spandrels
are attached to each floor level instead. The system is very economical and a
common facade type. To resist air-blast loads, steel angle kicker braces are often
needed to laterally support the bottom of the spandrel panel.
Glass Curtain Wall: For this glazing type, a significant part of the building exterior is covered with windows supported by aluminum or possibly steel framing.
Figure 10.5 shows three variations of curtain wall system.
Figure 10.4 Example of Ribbon Windows, Showing Spandrel
(a)
(b)
(c)
Figure 10.5 Types of Metal and Glass Curtain Walls. (a) Curtain wall framing expressing main building structure. (b) Uniform curtain wall framing over the fac¸ade. (c) curtain
wall framing draped over rectilinear building forms.
FENESTRATION
277
Figure 10.6 Example of Point and Cable Supported Curtain Wall
This support type has an increased risk of hazardous failure as compared with
punched windows because strip window systems are supported by wall on two
sides instead of four sides.. However, it has more flexibility, which allows it to
deform significantly without failing.
Point and Cable Supported: In point and cable supported systems (see
Figure 10.6), each window pane is connected at points near its corners. The
bracket that supports the glass often has multiple arms, forming a spider, so that
it can attach near the corner of several adjacent panes. The glass is drilled for a
threaded connector to pass through, engaging the interlayer to secure the pane
during an explosion. The panes are typically connected using a clear or translucent polymer material instead of the metal framing (referred to as “butt glazed”).
Spandrels The term “spandrel” is used to describe the non-vision panels used
above and/or beneath the window or at the floor levels (see also discussion regarding “strip windows” above). They are used to hide from view the structure
or mechanical equipment behind. Sometimes a “shadow box” is used behind the
spandrel, which is a metal pan designed to catch the glass fragments and keep
the explosive loads from entering the building.
Operable versus Inoperable Although operable windows can be designed to
meet modest explosion requirements, keep in mind that they will not keep air
blast out of the building if they are open.
Inoperable window solutions have potential to be more reliable for protection
during air-blast events, because occupants cannot void the windows’ protective
function by opening them. However, there are operable window solutions that
are conceptually viable. For instance, if a window is designed to open outward
278
BUILDING ENVELOPE AND GLAZING
about a horizontal hinge at the sill, the window will tend to slam shut in an
exterior explosion. If this type of design is used, the governing design parameter
may be the capacity of the hinges and/or hardware.
The controlling design component for sliding operable windows (and glazed
doors) often is the supporting track. If possible, the track should be embedded in
the supporting structure to provide added out-of-plane support to the system.
Skylights/Courtyards Skylights in roof systems have the advantage of not being subject to reflected pressures fromexterior explosions at ground level. Also,
skylights are required by building code to be designed using laminated glass,
which is preferred for explosion-mitigating design. However, they do create a
falling fragment hazard. Therefore, these should be designed with a catch system beneath, or designed to remain in the frame for the design air-blast load.
Ideally, skylights should be placed as far from the weapon as possible, to keep
the pressures low.
For an atrium with an open courtyard in the center, the windows are protected
by the building on all sides, and the glass will be subject to indirect air-blast
pressures in the event of an explosion outside the building exterior. Because atria
are protected on all sides, they may be designed for the reduced pressures from an
exterior threat. These reductions in pressures can help significantly in reducing
the design requirements imposed by explosion loads.
Doors Glazed doors and glass panes above a door are to be designed using
the same methods as for windows. Glazed doors tend to use tempered glass for
safety, in case of accidental breakage due to impact. As a result, they are highstrength and can complicate the design of the supporting system
10.3.2 Mullions/Transoms
The vertical frame members connecting adjoining windows are referred to as
mullions. Horizontal members are termed transoms. These members may be
designed in two ways. Either a static approach may be used, whereby the
breaking strength of the window glass is applied to the mullion, or a dynamic
load may be applied, using the peak pressure and impulse values. A static approach may be overly conservative because the designer using this approach
assumes that the peak edge forces are sustained indefinitely, whereas in an
explosion the forces are of very short duration, and the supporting structure
has insufficient time to fully respond to the impulsive forces. Using this approach, the mullion can become very deep and heavy, driving up the weight
and cost of the window system. Figure 10.7 shows heavy mullions in an older
building.
Sometimes cables or steel bars or tubes attached to the supporting structure are
placed behind the glass to prevent the laminated glass or glass with anti-shatter
film from entering the interior (Crawfordk, Lan and Dunn 2006).
FENESTRATION
279
Figure 10.7 Example of Mullions in an Older Building
10.3.3 Frame and Anchorage
The window frames need to retain the glass so that the entire pane does not
become a single large unit of flying debris.
To retain the glass in the frame, a minimum of a 1/4 -inch bead of structural
sealant (e.g., silicone) is used around the inner perimeter of the window. For
designs where a 60-mil polyvinyl butyral (PVB) interlayer is used, the silicone
sealant should be designed to resist the shear forces caused by the membrane action forces using the ultimate tensile capacity of the PVB material. These membrane forces can significantly increase the reaction loads into the framing system,
increasing the design requirements for the entire window system. The allowable
tensile strength of the silicone sealant should be at least 20 psi. Also, the window
bite (i.e., the depth of window captured by the frame) needs to be at least 1/2 -inch.
For large windows, such as in the lobby area, the bite required for conventional
loads may well exceed the 1/2 -inch minimum, and additional silicone sealant may
be needed to avoid failure caused by the entire glass pane exiting the frame.
280
BUILDING ENVELOPE AND GLAZING
Frame and anchorage design is performed by applying the reaction forces
from the window at breakage to the frame and the fasteners. In most conventionally designed buildings, the frames will be aluminum. In some applications,
where the windows are designed to resist high pressures and for long spans, steel
bar inserts, cable inserts, or built-up steel frames may be used.
For reinforced concrete construction designed to resist high-pressure loads,
as is typical for embassy construction, anchorage of the steel window frames is
provided by steel studs welded to a steel base plate. For this type of construction,
the frame is typically constructed using a steel stop at the interior face and an
angle with an exposed face at the exterior face. The frame is attached to the base
plate using high-strength fasteners. Coordination between the wall and window
subcontractors is required to ensure that the fastener locations are spaced so that
they fit between the rebar in the wall.
For masonry walls, flat metal straps embedded in the mortar with adequate
development length are recommended for anchoring the window into the wall.
10.3.4 Supporting Structure
It is inconsistent with balanced design and is potentially highly hazardous to have
a wall system that is weaker than the windows it is supporting.
Anchoring window/wall systems into columns is generally discouraged because it increases the tributary area of lateral load that is transferred into these
critical members, and may cause failure or instability.
Some window/wall designs will require additional lateral support. For instance, when the supporting walls are acting largely as cantilevers, such as for
windows placed high on a wall, they might need to be supported with vertical
braces spanning the clear floor height using for instance steel tubes to provide
additional stiffness. For punched wall systems with narrow pilasters between
them, vertical braces may also be needed. For lighter wall systems such as metal
stud systems, suitable reinforcements such as back-to-back double studs framing
the window are recommended.
The balanced design approach is particularly challenging in the design of
ballistic-resistant and forced-entry-resistant windows, which consist of one or
more inches of glass and polycarbonate. In this situation, the window may be
framed directly to the structure using steel framing to transfer loads and not into
the wall.
10.3.5 Other Penetrations
Similar to windows, other penetrations are designed to improve the safety of the
occupants. Rather than explicitly designing to resist air-blast pressures, the focus
is on reducing the hazard presented by the failure.
Doors Doors are handled differently in different criteria documents. Most criteria documents neglect the response of door systems. This may be for several
reasons. Doors that are capable of resisting air-blast loads can be very expensive.
EXTERIOR WALLS
281
Also, doors are typically in transitory areas where people do not stay for very
long. Some concepts for increasing the inherent protection offered by doors are
as follows:
r Use double steel doors with internal cross braces.
r Orient doors to open outward so that they bear against the jamb during the
positive pressure loading phase.
r Fill the jambs with concrete to increase their strength.
r Increase the number of fasteners used to connect the door into the wall
system.
Revolving doors are required by code to use tempered or laminated glass, to
prevent impact breakage during normal use. It is not typical to explicitly design
these doors to resist the threat, but some precautions may be taken to reduce the
hazard, such as increasing the bite or using silicone sealant between the glass
and frame.
Louvers Louvers are another type of opening to consider. These components
should be designed with connections that are able to resist the flexural capacity
of the louver. A catch system, consisting of a well-anchored steel grate behind
the louver, is another approach.
Air intakes that are at or close to the ground level should always have grates so
that weapons cannot be lobbed into them. Also consider using a sloped grating
(at an angle of at least 45 degrees from the horizontal) so that a potential weapon
can roll off prior to detonation.
Blowout Panels Blowout panels are designed to disintegrate or break free at
relatively low impulses, to limit the buildup of expanding hot gasses in an interior
space. This type of wall fails easily and will act to vent the explosive forces outside the building, helping to protect vertically and horizontally adjacent spaces.
Historically, blowout panels also are used when designing for an accidental
explosion inside a building. For deliberate explosions, the approach is somewhat different. Blowout panels typically are used when an internal space, such
as a mailroom designed to resist a package bomb, is vulnerable to explosive attack. In these situations, mailrooms are placed adjacent to the exterior wall of
the building, and the exterior wall is designed using lightweight construction using, for instance, metal studs. Alternatively, conventional construction, without
consideration to explosion loads, may be used.
10.4 EXTERIOR WALLS
Exterior walls often are designed to resist reflection-magnified pressures from
an external explosive threat located directly beyond the secured perimeter line.
One important objective of the design is to detail these members to fail in a
282
BUILDING ENVELOPE AND GLAZING
ductile mode such as flexure, rather than a brittle mode such as shear. The walls
also need to be able to resist the ultimate loads transmitted by the windows and
doors.
It may be cost-efficient to consider the reduction in pressure with height due to
the increase in distance and the angle of incidence at the upper levels of a highrise building. Even if pressure reductions are taken into account at the upper
floors, minimum requirements such as balanced design and ductile response are
to be met, to reduce the hazard to occupants in case the actual explosion is greater
than the design threat.
Various types of wall construction are considered below.
10.4.1 Concrete Walls
Cast-in-Place Historically, the preferred material for explosion-mitigating
construction is cast-in-place reinforced concrete. This is the material that is used
for military bunkers, and the military has performed extensive research and testing on its performance. Reinforced concrete has a number of attributes that make
it an attractive construction material because of its mass, ductility, and monolithic
character.
Reinforced concrete is recommended for high-risk buildings that are vulnerable to large-scale attack, and require a high level of protection. It is also recommended for highly protected areas within buildings, such as primary egress paths
or high-occupancy areas.
Note that for reinforced concrete to respond favorably to explosion loads, it is
to be detailed in a ductile manner, such as is done in high seismic zones. Some
attributes of ductile protective design (Federal Emergency Management Agency
2003) are as follows:
r Use symmetric reinforcement on both faces, so that load reversals may be
accommodated.
r Span the wall from floor to floor, rather than from column to column, to
reduce the chances of building collapse.
r Stagger splices away from high-stress areas, to maintain the monolithic
character of the material.
r Space reinforcing bars no more than one wall thickness apart, but no less
than one-half the wall thickness apart, to lessen the chances of localized
brittle behavior.
r Use ductile special seismic detailing at plastic hinge locations, so that they
are able to develop the full moment of the span.
r Use full tensile development lengths to develop the ultimate flexural capacity of the section.
r For progressive collapse prevention, consider the loss of an exterior wall
that measures vertically one floor height and laterally one bay width.
EXTERIOR WALLS
283
r Use closed ties or spiral reinforcing along the entire length of beams and
columns, including connections with a minimum bend angle of 135◦ and a
spacing not exceeding d/2, to provide confinement.
Precast Panels For precast panels, consider a minimum thickness of five
inches, exclusive of reveals, with two-way reinforcing bars to reduce the ultimate
resistance, increase ductility, and reduce the chance of flying concrete fragments.
This will also reduce the reaction loads transmitted into the connections. In highpressure regions, using ribbed panels (see Figure 10.7) may be an effective way
to resist the loads. These panels need to bear against the floor diaphragms.
The following are recommendations and considerations in designing precast
elements for air-blast resistance.
Reinforcement: Two symmetric, continuous layers of two-way reinforcement
are recommended to accommodate large deformations and rebound loads. For
thin panels where it is difficult to place two layers of reinforcement, the use of
two layers of heavy wire mesh, one layer of two-way reinforcement along the
centerline, or staggered bars on either face may be considered.
If a single layer of reinforcement is used, it is critical to design the section so
that the steel yields before the concrete fails in compression to obtain a ductile
response. Enhanced protection may be provided by placing fiber-reinforced polymers, geotextile materials, Kevlar, or similar materials on the inside face to provide confinement, resistance to scabbing and spalling, and added tensile capacity.
If reinforcing bars are used, a layer of wire mesh on the interior face may help to
further restrain concrete fragments from entering the space. Closely spaced bars
also help fragment restraint, but care must be taken not to increase the ultimate
Figure 10.8 Precast Concrete Ribbed Wall Panels
284
BUILDING ENVELOPE AND GLAZING
flexural capacity too much, so as to keep the reaction loads to a reasonable level.
Note that centerline reinforcement will not work in the rebound direction due
to the failure of the tension concrete during the positive phase. If the primary
objective is to protect occupants and not passersby, then this may be acceptable.
Above major egress points, however, where debris outside the building presents
an obstacle to ingress and egress post-event, added protection is desirable.
Load-Bearing Systems: For progressive collapse resistance in load-bearing
precast systems, panels need to be designed to span over failed areas by means
of arching action, strengthened gravity connections, secondary support systems,
or other means of providing an alternate load path.
Connections: Ductile connections are recommended. The connections and
supporting structure need to be able to resist the loads transmitted by the panel
loaded to its ultimate flexural capacity, so that the system is balanced for the possibility that the actual air-blast loading is higher than the design load. Reaction
loads from the windows at ultimate capacity are to be included in the calculation
of connection design loads. Using this approach, every panel with a different
configuration will have a different set of design loads for the connections. Note
that small panels will have higher reaction loads than larger panels, using this
method, due to their increased stiffness. Standardizing the panel sizes greatly
simplifies the connection load determination.
Also, the connections need to provide sufficient lateral restraint for the panels
to accept large deformations. Depending on the design details of the connection
used, lateral restraint design will require consideration of in-plane shear, buckling, flexure, and/or tension loads in the design. Punching shear through the panel
also needs to be checked. The connection should provide a direct load path from
the panel into the supporting structure to minimize P-Delta effects.
Floor-to-floor panels with continuous, gravity-type connections directly into
the floor diaphragms are preferred. Multistory panels that are hung on the floor
diaphragms may also be considered. Connections into exterior columns or spandrel beams are discouraged, to avoid the possibility of initiating structural collapse of the exterior bay. Redundant gravity connections are strongly recommended, to prevent falling debris if a single connection fails.
Connections should be checked for rebound loads. It is conservative to use
the same load in rebound as for the inward pressure. More accurate values may
be obtained through dynamic analysis or charts provided in military handbooks
(Unified Facilities Criteria Program 2008).
Specifications: Specifications for precast elements should be in the form of
a performance requirement, with the air-blast pressures and required response
limits. The performance specifications give precast contractors more flexibility
to provide the systems with which they are most familiar. This approach requires
that the contractors have either in-house dynamic analysis capability or a relationship with a blast engineer who can work with them to customize the most
cost-effective system.
For structural precast systems, the connections are the critical issue to be
addressed. Connections need to permit the panel to reach its ultimate flexural
EXTERIOR WALLS
285
capacity. Extensive use of transverse walls or an “egg crate” type of design is
an effective way to achieve the needed lateral support. Special seismic detailing (American Concrete Institute 2008) is recommended for structural precast
connections.
Tilt-Up Construction Tilt-up construction is acceptable, provided that the connections into the structure are able to withstand the ultimate capacity of the wall
in both the inward and outward (i.e., rebound) direction.
Post-Tensioned Panels Pre-tensioned or post-tensioned construction without
significant mild steel reinforcement provides little capacity for abnormal loading
patterns and load reversals. If these systems are used, it is recommended that
mild reinforcing be added, to provide the needed ductility
10.4.2 Masonry
Reinforced Masonry For concrete masonry unit (CMU) block walls, as a minimum use 8-inch block walls, fully grouted with vertical centered reinforcing
bars placed in each void, and horizontal ladder type reinforcement at each layer.
Connections into the structure are to resist the reactions associated with the wall
loaded to its ultimate lateral capacity. For infill walls, avoid transferring loads
into the columns if they are primary load-carrying elements.
The connection details may be very difficult to construct. It will be difficult to
have all the blocks fit over the bars, near the top, and it will be difficult to provide
the required lateral restraint at the top connection. A preferred system is to have
a continuous exterior CMU wall that laterally bears against the floor system.
For infill walls, use development lengths that develop the full capacity of the
section. For increased protection, consider using 12-inch blocks with two layers
of vertical reinforcement. Also, consider using CMU block units that encourage
a homogenous response, such as an “I”-shaped unit with inner and outer faces
that are connected with a small strut.
Unreinforced Masonry Brick load-bearing walls are unable to sustain tensile
loads, so they must resist explosion loads mostly through mass to reduce the
kinetic energy that must be absorbed. As a result, thicker solid walls on the
order of approximately 18 inches are preferred and respond well at air-blast levels less than 10 psi or so. Unreinforced structural masonry is considered a very
brittle material that may generate highly hazardous flying debris in the event of
an explosion, and is to be avoided for new construction.
10.4.3 Steel
Metal Studs For metal stud systems, use metal studs back to back and mechanically attached. The two-stud system provides the benefit of lateral torsional resistance, and therefore can be more efficiently designed. To catch exterior cladding
286
BUILDING ENVELOPE AND GLAZING
fragments, attach a wire mesh, steel sheet, or fiber-reinforced polymer to the interior side of the panel. The supports of the wall should be designed to resist the
ultimate out-of-plane bending capacity load of the system.
If a single-stud system is used, deeper, thicker channels are preferred. As a
minimum, consider 16-ga systems with depth of 6 inches, and laterally braced
are preferred.
Special care is required at the connections to prevent failure prior to the stud
reaching its ultimate capacity.
Enhanced protection may be provided by placing fiber-reinforced polymers,
geotextile materials, Kevlar, or similar materials on the inside-face of the
cladding to provide confinement, fragment restraint, and added tensile capacity.
Metal Panels Non-structural metal panels may be successfully used if they
are braced behind with steel tubes or compact sections. Continuous welds are
required around the perimeter braces to resist the membrane loads. At the end
panels, the braces will need to be augmented to account for the lack of equalizing
forces.
10.4.4 Other
Appurtenances Brick veneers, bris soleil, sunshades, and other nonstructural
elements attached to the building exterior are to be avoided, to limit flying debris
and improve emergency egress by ensuring that exits remain passable. If used,
they should be designed using lightweight materials with connections designed
to resist the ultimate capacity of the appurtenance.
Exposed Structural Systems For exposed exterior frame systems, there are two
primary considerations. The first is to design the exterior columns to resist the
direct effects of the specified threats. The second is to detail the exterior frame so
that it has sufficient structural integrity to accept localized failure without initiating progressive collapse. As a rule of thumb to meet these goals, column spacing
should be limited to 30 feet, and floor heights should be limited to not greater
than 16 feet, wherever possible. It is recommended that the columns be designed
to resist the explosive loads associated with a contact charge placed at the base,
or that a column removal analysis be performed to verify that progressive collapse is not initiated.
Because columns do not have much surface area, air-blast loads on exposed
standalone columns that are not supporting adjacent wall systems tend to be
mitigated by “clearing-time effects” (Unified Facilities Criteria Program 2008).
This refers to the fact that the pressure wave washes around these slender tall
members, and consequently the entire duration of the pressure wave does not act
upon them. Figure 10.9 shows an exposed column support on the corner of a
high rise office building.
For columns subject to a vehicle weapon threat close by in the public street
in an urban area, buckling and shear are the primary effects to be considered in
EXTERIOR WALLS
287
Figure 10.9 Example of column freely exposed to public access. The column supports
the corner of a high-rise building.
analysis. This is a common scenario considered for office buildings in urban areas. Because exposed columns are often in public areas and are primary support
members for buildings, the concern is that progressive collapse may be initiated.
A very large weapon close to a column can cause shattering of the concrete
due to multiple tensile reflections within the concrete section, destroying its integrity. Circular columns shed load more rapidly than rectangular columns and
can be beneficial.
Buckling is a concern if lateral support may be lost when a floor system that
provides lateral support is damaged. This is a situation that arises frequently for
office buildings that are close to public streets. In this case, exterior columns
should be capable of spanning two or more stories without buckling.
Insetting the first line of columns a few feet into the building interior is a
nonconservative approach to protecting the columns. Even if the columns are
not publicly accessible, they are still vulnerable to air-blast effects, and their
response still needs to be addressed.
Protect columns by using closely spaced closed ties or spiral reinforcing, to
improve confinement and shear capacity. This also will improve the performance
of lap splices in the event of loss of concrete cover, and greatly enhance column
ductility. The potential cost-to-benefit ratio for providing closely spaced closed
ties in exterior concrete columns is among the lowest and should be considered
seriously. Closed ties are to be used in columns and spandrels for the entire span
and across connections with a maximum spacing of d/2 and a minimum bend
angle of 135◦ . Some other recommendations for reinforced concrete are given in
wall subsection 10.4.1.
For exposed steel columns, splices should be placed as far above grade level
as practical, and base plates should be recessed below the ground level.
288
BUILDING ENVELOPE AND GLAZING
For a contact or close-in package weapon, column breach is a major consideration. To mitigate this threat, some suggestions include:
r Do not use exposed columns that are fully or partially accessible from the
building exterior. Arcade columns should be avoided (see Figure 10.10).
r Use an architectural covering that is at least 6 inches from the structural
member. This will make it considerably more difficult to place a weapon
directly against the structure. Because explosive pressures decay so rapidly,
every inch of distance will help to protect the column.
r Use steel plates to surround the base of concrete columns where it is accessible. Plates need to extend several feet above the accessible location.
r Encase steel columns with concrete and add a plate, 3/8 –1/2 thick around
the perimeter for about 5 feet above the base, if necessary.
Load-bearing ductile reinforced concrete wall construction without columns
can provide a considerable level of protection if adequate reinforcement is provided to achieve ductile behavior. This may be an appropriate solution for the
parts of the building that are closest to the secured perimeter line (at scaled ranges
less than 3).
Spandrel beams of limited depth generally do well when subject to air blast.
In general, edge beams are very strongly encouraged at the perimeter of concrete
slab construction, to afford frame action for redistribution of vertical loads and
to enhance the shear connection of floors to columns. Confinement of concrete
spandrels using closely spaced closed ties, such as those used for columns, is
recommended. Transfer trusses and transfer girders are to be avoided because
they increase susceptibility to progressive collapse.
Figure 10.10 Example of public walkway beneath second floor of high-rise building
ROOF SYSTEMS
289
10.5 ROOF SYSTEMS
The primary loading on the roof is the downward air-blast pressure. The exterior bay roof system on the side(s) facing an exterior threat is the most critical.
Roof systems that are low and therefore closer to the explosion will be subject to
higher pressures than high roof systems. The air-blast pressures on the interior
bays are less intense and may require less hardening. Secondary loads include
upward pressure due to the air blast penetrating through openings, and upward
suction during the negative loading phase. The upward internal pressures may
have an increased duration due to multiple reflections of the internal air-blast
wave. It is conservative and recommended to consider the downward and upward loads separately.
Because roof systems are not exposed to the reflected wave from the ground
level, they are subject to lower pressures than the walls facing the explosion.
Because the angle between a sloped roof and the advancing shock sometimes
is less acute than the angle between a flat roof and the advancing shock front,
sloping roofs may be subject to somewhat higher pressures. They are usually of
lighter construction than flat roof systems, making them particularly vulnerable
to air-blast effects unless designed to respond in a ductile manner. For this reason,
sloped systems sometimes require more robust designs.
10.5.1 Concrete
The preferred system is to use cast-in-place ductile reinforced concrete with
beams in two directions. If this system is used, beams should have continuous,
symmetrical top and bottom reinforcement with tension lap splices. Shear ties
should develop the bending capacity of the beams and be closely placed along
the entire span. All ties are to have a 135-degree bend minimum. Two-way slabs
are preferred.
Precast and pre/post-tensioned systems, including hollow plank, are generally
viewed as less desirable due to the lack of ductility. If they are used, a system
that has continuous bond with the concrete is preferred, with anchors that are
designed to be protected from direct air-blast effects. Also, additional mild reinforcement at top and bottom is recommended to ensure a ductile response.
Connections need to be designed to resist both the direct and uplift forces.
Concrete flat slab/plate systems are also less desirable because of the potential
of punching shear failure at the columns. Where flat slab/plate systems are employed, they should include features to enhance their punching shear resistance.
Continuous bottom reinforcement should be provided through columns in two
directions, to retain the slab in the event that punching shear failure occurs. Edge
beams should be provided at the building exterior.
10.5.2 Steel
Lightweight systems, such as untopped steel deck, are considered to afford negligible resistance to air blast. These systems are prone to failure due to their low
capacity for downward and uplift pressure.
290
BUILDING ENVELOPE AND GLAZING
10.5.3 Composite
Modest levels of protection are afforded by conventional steel beam construction with a steel deck and concrete fill slab. The performance of this system can
be enhanced by using normal-weight concrete fill, increasing the gauge of deck
and welded wire fabric reinforcement from that required for conventional loads,
and making the connection between the slab and beams using additional shear
connector studs. Tension membrane behavior along the edges should be considered in the design of the connections by using welding or fasteners along the
edges. Since it is anticipated that the slab capacity will exceed that of the supporting beams, beam end connections and supporting columns should be capable
of developing the ultimate flexural capacity of the beams, to avoid brittle failure.
Beam to column connections should be capable of resisting upward as well as
downward forces.
10.5.4 Penthouses/Gardens
Parapets, roof mechanical room enclosures, and tile roof systems are generally
not a primary concern, since they are exterior to the building. Generally these
members are designed to sustain heavy damage but not become flying debris.
Although roofing aggregate may become a flying hazard, in the context of an
explosion event, this hazard is not significant enough to warrant much concern.
Soil can be highly effective in reducing the impact of a major explosive attack.
Bermed walls and buried rooftops have been found to be highly effective for military applications and can be effectively extended to conventional construction.
This type of solution can also be effective in improving the energy efficiency of
the building.
10.6 BELOW GRADE
For buildings that are very close to the secured perimeter, there is the possibility
of the foundations becoming undermined by the cratering effects of an explosion.
However, if this is an issue, then generally, it will be accompanied by heavy damages to the superstructure as well. If the crater is projected to reach the building,
then the most cost-effective option may be to increase the building setback.
Ground shock effects are generally a secondary effect since most of the energy
of a vehicle weapon is transmitted to the air rather than the soil. The weapon
would have to be placed underground to have a significant effect on the structure.
Currently, underground weapons are not considered by the governing federal
criteria for civilian buildings.
There are significant benefits to placing secured areas below grade in terms of
mitigating explosion effects from an exterior weapon. The massiveness and softness of the soil provide a protective layer that significantly reduces the impulse
on the structural systems below grade.
BELOW GRADE
291
For buildings with below-grade portions that are adjacent to the building,
when creating a plaza level at ground level, keep in mind that the roof systems of
these underground areas will need to be designed for the actual air-blast pressure
levels, if these are occupied areas. If these are unsecured areas, such as a garage,
consider letting the roof fail if adequate egress routes are available on other sides
of the building, away from the failed plaza level.
Another consideration for belowground portions of the building is the design
of the perimeter security barriers. The perimeter barriers often require deep foundations, which may interfere with the underground structure.
It is preferable to place underground garages adjacent to the main structure
rather than directly underneath the building, to protect the structure against the
effects of an internal weapon. The effects of an internal weapon are generally
not a major concern for foundation walls. The soil on the other side of the wall
provides a buffer that mitigates the response. One exception to this is a situation
where the foundation is below the water table, where even a localized breach of
the wall may cause extensive collateral damage.
Vaults for transformers placed beneath the ground close to public streets are
of concern. If possible, place these away from public streets. If they are in a
driveway, the vault lid needs to be designed to resist the downward pressure.
In high seismic regions, seismic isolators may be used at the base of a building
(Figure 10.11) . In this case, the response of the building globally should be
checked for the total air-blast loading acting on the side facing the explosion.
Preliminary studies investigating this issue have shown that seismic response
governs.
moat cover
ground floor
ground level
moat
retaining wall
flexible utility connection
basement
base isolator
Figure 10.11 Diagram of seismic base isolation system. The building superstructure is
detached from the foundation and supported by steel plat/rubber bearings.
292
BUILDING ENVELOPE AND GLAZING
10.7 REDUCTION OF BLAST PRESSURES
The placement of the building on the site can have a major impact on its vulnerability. Ideally, the building is placed as far from the property lines as possible.
This applies not only to the sides that are adjacent to streets, but to the sides that
are adjacent to adjoining properties as well, since we cannot be certain about
how access will change for those neighboring properties during the life of the
building. A common-practice example of this is the use of a large plaza area in
front of the building, which often leaves little setback on the sides and rear of the
building. This practice can diminish the vulnerability of the front of the building,
but generally will increase the vulnerability of the other three sides.
The shape of the building can have a contributing effect on the overall damage
to the exterior envelope. Reentrant corners and overhangs are likely to cause
multiple reflections of the air blast, which may amplify the effect of the air blast
(see Figure 10.12).
(a)
(b)
Figure 10.12 Reentrant corner plans: (a) Some plan types. (b) T-shaped building with
two reentrant corners.
REDUCTION OF BLAST PRESSURES
293
In general, convex rather than concave shapes are preferred for the exterior of
the building (i.e., the shock front incidence angle on a convex surface increases
more rapidly with lateral distance from a detonation location than on a planar surface, causing the reflected pressure on the surface of a circular building to decay
more rapidly than on a flat building, See Figure 10.13). Similarly, the air-blast
pressures decay with height, as the angle of incidence becomes more oblique.
The sides of the building not facing toward the explosion do not experience reflected pressure and will typically perform with less damage.
(a)
(b)
Figure 10.13 Building shapes. (a) Convex, and (b) Concave.
294
BUILDING ENVELOPE AND GLAZING
Generally, simple geometries, with minimal ornamentation (which may become flying debris during an explosion), are recommended unless advanced
structural analysis techniques are used. If ornamentation is used, it is recommended that it consist of a lightweight material, such as timber or plastic, which
is less likely than brick, stone, or metal to become lethal projectiles.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
and Commentary (ACI 318-08), Chapter 21. Farmington Hills, MI: American Concrete
Institute.
Amstock, Joseph S. 1997. Handbook of Glass in Construction. New York: McGraw-Hill.
ASTM. 2001. Standard Test Method for Structural Performance of Glass in Exterior
Windows, Curtain Walls, and Doors Under the Influence of Uniform Static Loads by
Destructive Methods (E997-01). West Conshohocken, PA: ASTM International.
. 2003. Standard Practice for Specifying an Equivalent 3-Second Duration Design Load for Blast Resistant Glazing Fabricated with Laminated Glass (F2248-03).
West Conshohocken, PA: ASTM International.
. 2004a. Standard Practices for Determining Load Resistance of Glass in Buildings (E1300-04e1). West Conshohocken, PA: ASTM International.
. 2004b. Standard Test Method for Glazing and Glazing Systems Subject to
Airblast Loads (F1642-04). West Conshohocken, PA: ASTM International.
Computers and Structures, Inc. 1990. SAP2000/NonLinear User’s Manual. Berkeley,
CA: Computers and Structures Inc.
Crawford, John, Shengrui Lan and Brian W. Dunn. 2006. Cable catcher systems for improving blast resistance of glazing facades. In Proceedings of the 19th Military Aspects of Blast and Shock Symposium, Alberta, Canada, October 2006. Burbank, CA:
Karagozian & Case.
Federal Emergency Management Agency. 2003. Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks (FEMA 427). Washington, DC: Risk Management
Series of Publications, FEMA, Department of Homeland Security.
Hyde, David W. 1988. Microcomputer Programs CONWEP and FUNPRO, Applications
of TM 5-855-1, Fundamentals of Protective Design for Conventional Weapons (User’s
Guide). Vicksburg, MS: Army Engineer Waterways Experiment Station, Structures
Lab.
Livermore Software Technology Corporation. 1999. LS-DYNA User’s Manual, Nonlinear Dynamic Analysis of Structures, Version 950. Livermore, CA: Livermore Software
Technology Corporation.
MathSoft. 2001. Mathcad User’s Guide. Cambridge, MA: MathSoft Inc.
MathWorks. 2006. MATLAB: The Language of Technical Computing, Version 7. Natick,
MA: The MathWorks, Inc.
Unified Facilities Criteria Program. 2007. DoD Minimum Antiterrorism Force Protection
Standards (UFC 4-010-01) Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
REDUCTION OF BLAST PRESSURES
295
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Army Engineer Research and Development Center (ERDC). 2001. Window Fragment Hazard Level Analysis, the HAZL Model. Vicksburg, MS: U.S. Army Engineer
Research and Development Center.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
11
Protection of Spaces
MeeLing Moy and Andrew Hart
Security can be integrated at the earliest stage of planning and continued throughout the design and construction phases for building structures requiring blast resistance, or other specialized security measures. Building occupant emergency
plans may also be included in planning such that the design created can supplement the objectives of emergency operations. The goal is to increase the chances
that emergency systems will remain operational during an explosion. The continued operation of particular building elements during emergencies is essential
for building occupant life safety and evacuation. This chapter will focus on the
following building components: areas isolating interior threats, stairwell enclosures, hardened plenums, and safe havens.
11.1 AREAS ISOLATING INTERIOR THREATS
Interior explosive threats may be carried or delivered into accessible areas within
a building. These areas include, but are not limited to, loading docks, mailrooms,
and lobbies. The magnitude of interior threats and the specific areas isolating
them within a building are identified by a risk assessment. The Federal Emergency Management Agency (FEMA) defines a small explosive to be approximately 5 to 10 pounds of TNT equivalent, a large explosive to be approximately
50 to 100 pounds of TNT equivalent, and a mail bomb to usually be less than 10
pounds of TNT equivalent.
Protection of critical building components and systems within the identified
interior high-risk areas may be achieved by installing electronic surveillance
cameras, by implementing controlled access with entry and/or exit checkpoint
screening, and by structural hardening.
Walls and slabs enclosing specific interior high-risk areas should be designed
or hardened for the identified interior threats as determined by a risk assessment.
FEMA recommends that blast effects should be vented through the walls isolating interior threats. For example, these walls may consist of blowout panels
designed to provide security from blast loads applied on the outside due to an
exterior threat, but to fail due to an interior threat, thus venting the blast loads
297
298
PROTECTION OF SPACES
outside. FEMA suggests that slabs isolating interior threats should be designed
or hardened to consider downward and upward blast loads.
11.2 STAIRWELL ENCLOSURES
Exit and emergency egress stairwells are considered to be spaces that are accessible to the general public. They are designed for optimal performance with sufficient distribution and redundancy to meet current building code requirements,
and building occupant needs. Stairwell planning and design are a fundamental
part of a building occupant emergency plan. FEMA recommends that stairwells
should maintain positive pressure for safe evacuation of building occupants, and
for access by firefighters and rescue responders during an emergency. The installation of special filtering systems to provide a clean source of air may be required,
to maintain positive pressure and minimize smoke and hazardous gases in stairwells. The National Institute of Standards and Technology (NIST) recommends
that stairwell design capacities should also take into account the counterflow of
rescue personnel during an emergency. Stairwell lighting should remain functional during an emergency, and stairwell exit signs should provide a clear and
continuous route to the outside or to another area that is safe. In addition to the
building code requirements, stairwells should be located in less vulnerable areas,
and far from areas where blast events may occur in the building. Thus, stairwells
should not discharge into high-risk areas.
Exit and emergency egress stairwell enclosures, which include walls and
slabs, should be designed or hardened for the identified threats specified in a
risk assessment. FEMA recommends that stairwell walls should be anchored to
the floor slabs at each end, in order to adequately transfer the loads from a blast
to the lateral resisting structural system of the building. FEMA suggests that
stairwell slabs should take into account the uplift pressures from a blast by using an upward load equal to the dead load plus half the live load for the floor
system. FEMA also recommends that doors to the stairwell enclosures should
consist of steel, and be adequately anchored to the stairwell walls. FEMA also
suggests that windows in the stairwell enclosures consist of laminated glass with
a 60-mil polyvinyl butyral (PVB) interlayer. The laminated glass should be adhered within the window frames and mullions using a 1/2-inch bead of structural
silicone. The window frames and mullions should be anchored to the stairwell
walls in order to develop the full capacity of the window glazing.
11.3 HARDENED PLENUMS
Mechanical equipment and systems are considered to be primary nonstructural
elements that are necessary for the life-safety procedures of a building. Their
configurations and capacities are designed with adequate redundancy to meet
current building code requirements and building occupant needs. They are
SAFE HAVENS
299
identified as critical assets that should continue main operations during an emergency, to ensure the safe evacuation of building occupants. FEMA recommends
that public access to critical mechanical equipment and systems should be
restricted, in order to reduce the potential damage from a blast event. In addition
to the building code requirements, primary and redundant mechanical equipment
and systems should be situated in separate locations and far from high-risk areas,
to reduce the possible hazard of both being damaged by a single blast event.
Structural elements supporting the plenum that surrounds the mechanical
equipment should be designed or hardened to resist reflected blast pressures
and impulses, such that debris and fragments are not propelled into the protected spaces. Mechanical vents and exposed plenums are designed or hardened
to the same level of protection as required for exterior walls. Designs should take
into account the potential infill of blast overpressures through mechanical openings. Structural elements and mechanical equipment mountings and attachments
should be designed to resist these internal pressures.
11.4 SAFE HAVENS
A safe haven can be defined as a secure area designed to protect occupants from
various hazards. Such shelters may be located outside or within a structure. Safe
havens are fully enclosed structures designed to resist the effects of blast loads,
and the impact of fragments and debris due to an explosive event. This section
will focus on safe havens as documented by FEMA.
11.4.1 FEMA Documents
The FEMA document that deals with the protection of safe havens is primarily FEMA 453, Safe Rooms and Shelters, Protecting People Against Terrorist
Attacks, dated May 2006. Within the document two types of shelters are defined:
r Standalone shelter
r Internal shelter
The difference between the two types of shelters is that a standalone shelter
is a defined as a building that is not attached to or within any other building; it
is considered as a separate building. This separate building is to be constructed
to withstand a range of natural or man-made hazards and may be situated away
from any debris or potential fragment hazards. The building is to be structurally
separated from any other building, and thus not be affected by any weakening
from an adjacent building collapsing. Since this is a separate building, it does
not need to be integrated into the building design, but it is costly compared with
an internal shelter.
An internal shelter is a specially designed room within or attached to a building. The room is designed to be structurally independent of the building and is
300
PROTECTION OF SPACES
able to withstand a range of natural or man-made hazards. Since the shelter is
partially shielded by the building, it may not get the full blast pressure. The location of the room within the building has to be easily accessible by all building
occupants, and when designing for blast mitigation, no resistance from any of
the surrounding areas should be included. Having it within the building is less
costly than building a standalone shelter, as it can be part of a renovation or a
later addition to the building design.
11.4.2 Multi-Hazard Threats
FEMA 453 does not quantify a specific threat such as a terrorist threat from a
vehicle bomb; rather, the document gives general guidelines on different types of
building construction and reasonable mitigating measures for providing a secure
shelter.
Prior Warning In some instances, prior to an event, a warning is given. All
types of warnings given about any type of threat should be treated in a similar manner. The standard procedure for a prior warning is laid out in FEMA
453: Once the warning has been received, the contacted person calmly asks for
the location of the weapon (e.g., the building where it is housed), the type of
weapon, and any other pertinent information that he or she can obtain. Once
obtained, the contacted person must immediately call his/her local enforcement
agencies—such as local and federal law agencies—and emergency services, and
pass on all information from the perpetrator. All warnings should be treated as
reliable, regardless of their apparent validity.
Sequential Events after/during an Event Every type of shelter, regardless of
its size and number of occupants, should have an Emergency Operations Center,
or EOC. This center or group of people serves as an information management
center where decisions are made, such as those made during the attacks on the
twin towers on September 11, 2001. Each EOC should be equipped with communication equipment, reference materials such as medical books, log books to
note down what happened and when, and any other tools necessary to respond
quickly and appropriately to an emergency, such as a defibrillator and a first aid
kit. Within this center, the Emergency Management Group (EMG) makes decisions based upon the information provided by the Incident Commander (IC) and
other personnel. The relationship between the EMG and the emergency services
or Emergency Operations Group (EOG) is shown in Figure 11.1. Note that both
the EMG and the EOG have directors, each in constant contact with each other
throughout the duration of the event, coordinating the responses to the event as it
unfolds.
The use of an IC is vital, as this person is responsible for the front-line
management of the response to the incident, tactical planning and execution,
determining if help is required, and relaying requests for assistance, if required,
SAFE HAVENS
Emergency Operations Group
(EOG)
301
Emergency Management Group
(EMG)
Facility Manager
Emergency Director
Emergency Director
Safety Officer
Operations Officer
Affected Area
Manager/Supervisor
Coordinator Team
consisting of security,
environment,
maintenance, human
resources, planning,
logistics, and public
relations
Emergency Medical
Team
Emergency Services Team
(Fire/HazMat/Police)
Figure 11.1 Relationship between the EMG and the EOG
through the EOC. It is recommended that the IC be a member of management
with the ability to make critical time decisions and to act on them.
Subsequent Fires Because of the lack of fuel, fires from an explosion event
do not usually occur unless the weapon is an incendiary-type device or the event
has set off a chain of events that have given rise to a fire. In the event of an
incendiary bomb, whose sole purpose is to cause a fire, a fuel source close by is
required to sustain the fire. In the case of a chain reaction, the fire usually results
not directly from the explosion itself but from a different source of ignition, such
as a hot fragment igniting a burst gas pipe. Again, these fires, unless they have an
external fuel supply, are usually small and can be easily extinguished by either
cutting off the fuel supply or suffocating the flame.
11.4.3 Design Requirements for Protective Shelters
As in any design, protective shelters need to follow the basic construction codes:
American Concrete Institute (ACI) 318 for concrete (American Concrete Institute 2008), American Institute for Steel Construction (AISC) 360-05 for steel
construction (American Institute for Steel Construction 2005).
It is advisable, before beginning construction of a protective shelter, that a preliminary threat analysis be performed, by the designer, potential shelter owner,
or other responsible party, utilizing the Building Vulnerability Assessment
Checklist included in FEMA 426, Reference Manual to Mitigate Potential
302
PROTECTION OF SPACES
Terrorist Attacks Against Buildings (Federal Emergency Management Agency
2003).
When the preliminary threat analysis has been performed, an independent design professional should perform a more thorough assessment of the shelter to
either confirm the preliminary assessment or modify it accordingly if any deficiencies are found within the design/assessment.
If an existing building is to be used as a shelter, an assessment utilizing the
FEMA 426 Checklist will help to determine the building vulnerabilities and thus
aid in its modifications. It should be noted that FEMA 426 provides guidance to
the building construction industry in how to reduce physical damage to buildings.
The purpose of the manual is to show various approaches that can be taken to
minimize the effects of terrorist threats.
Resistance to Progressive Collapse Progressive collapse occurs when a localized failure occurs and the adjoining members are overloaded. This causes
failure of the members, and a cascade or “pancaking” failure effect occurs. This
failure may result in a partial or total failure of the structure. The initial localized failure could be the result of a small parcel bomb in direct contact with a
critical structural element, or the source could be a vehicle-borne explosive device located a relatively short distance away. A larger explosive device located a
longer distance away is not likely to cause a single member to fail, but will probably affect a number of structural elements instead, which is not deemed failure
by progressive collapse. In addition, according to FEMA 453, Safe Rooms And
Shelters: Protecting People Against Terrorist Attacks (Federal Emergency Management Agency 2006), progressive collapse is not an issue for buildings with
three stories or fewer.
The use of transfer girders and nonductile, nonredundant construction is prohibited in the building construction, as it may give rise to structural systems
that are not tolerant of localized damage. The columns that support transfer
girders—and the transfer girders themselves—may be critical to the stability of
a large floor area. For interior shelters, the building surrounding the shelter must
be sufficiently hardened so that progressive collapse is not an issue. For these
buildings, it is required that the structural integrity of the building must still be
intact if a member fails.
Some of the characteristics that a building containing a shelter must have to
resist progressive collapse are:
r Mass: The larger the inertial resistance, the more resilient it is to failure.
r Shear Capacity: Connections and members need to be designed to prevent
failure.
r Capacity for Load Reversals: Some members may undergo several large
deformations before damping down. These load reversals can be in the form
of uplift pressures, which defy the normal convention of gravity load design.
Structural elements subject to blast effects, therefore, should be designed
for load reversal. For design purposes, it is usualto provide equal areas of
SAFE HAVENS
303
reinforcing steel on both sides of reinforced concrete members, to provide
the necessary resistance. The connections also need to be designed for load
reversals.
r Redundancy: The use of an alternative load path in the vertical load-carrying
system. This allows the gravitational loads to redistribute in the event of a
structural member failure.
r Ties: The use of an integrated system of ties can serve to redistribute the
loads. The ties need to be located perpendicular to the principal line of
structural framing.
r Ductility: The ability of a structural member or connection to have a large
deformation while still maintaining strength. Allowing the member to inelastically deform (e.g., plastically deform) causes large amounts of blast
energy to dissipate; however, special detailing is required.
Using the above guidelines for designing buildings with shelters, or shelters themselves, robustness is greatly improved and progressive collapse may
be avoided. For existing structures that are being modified to accommodate a
shelter, it may be possible to allow only one floor adjacent to the failed column
to fail, but if the structural members are retrofitted to develop catenary behavior,
the adjoining bays must be upgraded to resist the lateral forces. This may result
in more extensive retrofitting than is feasible or desirable. When such cases occur, isolate this region and risk the collapse of the adjoining bays by upgrading
the vulnerable localized columns. When upgrading these columns, a balanced
design approach must be obtained. It should be noted that upgrading an existing
building for an alternative load path system is potentially counterproductive.
For buildings that can provide a continuous load path that will support all
vertical and lateral loads, all structural components and fasteners used in the
connection system must be able to develop the full capacity of the member. In
order for this to take place, the capacity of each component must be balanced
with the capacity of the other members and their connections. As all applied
loads must be transferred eventually to the foundation/ground, the load path must
be continuous from the top structural member to the ground.
Location within the Building One of the most important factors when dealing
with a shelter is its location, including determining the best location for saving
lives. The location of the shelter will depend on how many occupants the shelter
has to hold. In addition to the effects of blast, the location should also consider
effects of chemical, biological, and radiological threats. The shelter should be
located such that all persons who are to take refuge can reach it within a minimal
travel time from where they are located. A way to ease travel time is to clearly
mark the route to the shelter.
Standalone shelters should be located as far as deemed possible from the surrounding buildings, to avoid both progressive collapse and/or impacts from the
collapse of nearby buildings. Internal shelters require that the main building be
designed for progressive collapse.
304
PROTECTION OF SPACES
Access to Egress Shelter locations should be such that all designated persons may take refuge within the shelter with a minimum of travel time. Travel
routes to the shelter should be clearly marked, allowing easy access. Exit routes
from the shelter should direct the persons away from the threat, and hazards
signs should be located using Crime Prevention through Environmental Design
(CPTED) principles. These principles make use of natural access control, natural
surveillance, and territoriality, which can also help to deter the aggressor.
Access to the building must be as “natural” as possible, in that the siting of
the entrances, exits, fences, lighting, and landscaping tends to guide people who
are entering or exiting the building. This is called “natural access control,” and
when it is coupled with limited access, the opportunity for a terrorist attack is
greatly reduced. The placement of external features such as bushes, trees, and
fountains must also be considered so that visibility is maximized. This is termed
“natural surveillance,” and it can reduce the opportunity for terrorists to attack,
by revealing their intentions; in addition, it gives a feeling of safety, as the public
can be easily seen. Finally, the use of fences, bollards, signage, and the like,
termed “territoriality,” clearly defines the property lines between what is public
and private. This can be a gateway into a community or neighborhood.
Fire Rating There are only two fire ratings, H and A (Steel Construction Institute 1990). The H fire rating is solely for hydrocarbon fires such as those caused
by a petroleum product. The A rating is for normal fires such as those that have
wood or paper as their source of fuel. A number usually follows the letter, such
as 60 or 120, and the value represents the time that structural integrity can be
maintained during a fire—e.g., how long a member or area can be kept below a
designated temperature, usually 40◦ C, before the temperature rises, causing that
member to plasticize and fail. In H-rated fires, the maximum temperature is designated as 1100◦ C within 10 minutes, and for A, the maximum temperature is
950◦ C, over 60 minutes (Steel Construction Institute 1990). See Figure 11.2. It
is usual for commercial buildings to be designated as A-rated, as the chances of
a hydrocarbon fire within one of these buildings is remote.
Currently, the fire protection of buildings, based on FEMA, is based on a
four-level hierarchy that comprises alarm and detection, suppression, compartmentation, and passive protection.
r Fire alarms and detectors are usually used to activate the system and notify
the building occupants and emergency services through verbal and visible
systems.
r Suppression systems, usually in the form of water sprinkler systems, are
primarily designed to control small to medium fires and to prevent the fire
spreading beyond the water zone, which is typically 1,500 square feet.
r Use of compartmentation will mitigate the spread of more severe fires.
These typically require fire-rated partitions between the rooms and can
cover an area of 12,000 square feet.
REFERENCES
305
Temperature Curves
1200
Temperature (C)
1000
800
A Rated
H Rated
600
400
200
0
0
5
10 15 20 25 30 35 40 45
Time (min)
50 55
60
Figure 11.2 Temperature profile of H- and A- Rated Fires
r Passive protection is the use of a protective covering on a steel structure.
This protection system will ensure that the temperature of the steel system
does not reach an undesirable level such that the steel becomes plasticized
and cannot withstand the vertical loading applied to it without buckling.
There are numerous different types of ratings for these types of products.
The purpose of this system is to prevent structural failure of the building
before the rated time of the product has elapsed, by which time the building
can be safely evacuated.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
and Commentary (ACI 318). Farmington Hills, MI: American Concrete Institute.
American Institute of Steel Construction. 2005. Specification for Structural Steel Construction (AISC 360-05). Chicago, IL: American Institute of Steel Construction.
Federal Emergency Management Agency. 2003. Reference Manual to Mitigate Potential
Terrorist Attacks against Buildings (FEMA 426). Washington, DC: Federal Emergency
Management Agency, Department of Homeland Security.
. 2006. Safe Rooms and Shelters: Protecting People Against Terrorist Attacks
(FEMA 453). Washington, DC: Federal Emergency Management Agency, Department
of Homeland Security.
Steel Construction Institute. 1990. Interim Guidance Notes for the Design and Protection
of Topside Structures against Explosion and Fire (SCI P-112). Silwood Park, Ascot,
Berkshire, UK: Steel Construction Institute.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
12
Defended Perimeter
Joseph L. Smith and Charles C. Ellison
The four main functions of any comprehensive physical security program are
to deter, detect, delay, and respond to a threat. A defended perimeter serves
an important role in achieving these aims. A well-planned perimeter can deter
an aggressor by increasing the perceived difficulty of attack, can support the
implementation and function of detection sensors, and can delay an aggressor,
providing the security force time to respond to an attack.
12.1 GOALS
The perimeter barrier system will generally be the outermost zone in a tiered
system of protections that includes the building facade and internal screening
zones such as visitor lobbies, mailrooms, and loading docks. These zones, as
shown in Figure 12.1, can be viewed as concentric layers of protection.
In terms of the primary topic of this book, the goal for each tier should be to
reduce the size of the likely explosive threat to a level that can be mitigated by
reasonable levels of blast hardening. As the first line of defense, the perimeter is
the foundation on which all other blast-mitigation options are selected. As such,
the defended perimeter must address the largest threats to the facility. Failure
of the perimeter to perform as expected can have severe consequences for the
adequacy of subsequent blast-mitigation methods.
Heavy barriers are not the only option when creating a defended perimeter.
Perimeter protection may include effective site planning and landscape design
that, if deployed properly, can substantially decrease the risk to occupants and
mitigate damage to a facility. The effective use of landscaping and terrain can
also minimize the design requirements and cost of other man-made barriers employed to stop vehicles or other threats.
12.2 STANDOFF
The standoff distance is generally considered to be the range between the center
of the explosive device and the nearest structural element requiring protection
307
308
DEFENDED PERIMETER
Figure 12.1 Zones of Protection
(i.e., the target). The term “setback” is used when discussing the distance between a facility and the defended site perimeter. These terms are often used interchangeably; however, standoff and setback do not necessarily have to be equal.
Since blast loads decay rapidly with distance, setback can be an effective
means of protecting a facility from an explosive attack. The first few feet provide
the most benefit. Figure 12.2 illustrates peak pressure versus standoff curves for
a range of explosive devices [Based on Equations in UFC 3-340-02, Structures to
Resist the Effects of Accidental Explosions (Unified Facilities Criteria Program
2005b)].
As shown in Figure 12.2, at very small standoff distances (i.e., less than 5 ft)
just an extra foot of additional standoff can reduce the blast load on a structural
member by an order of magnitude (i.e., a factor of 10) or more. It may only take
a few pounds of explosive in direct contact with a structural column or wall to
cause devastating structural collapse. However, that same column or wall may
resist several hundred pounds of explosives if only a few feet of standoff are
provided. The benefits of a few additional feet of standoff can also be significant
at much greater distances. For example, the pressure loads generated by a 500-lb
TNT bomb at 112 ft are over 20% less than the loads generated at 100 ft.
Even though the benefit per foot of standoff decreases with distance, providing standoff is often more cost-effective than structural hardening for significant
distances from the facility. To illustrate this point, the approximate relationship
STANDOFF
309
10,000
5 lb TNT
50 lb TNT
500 lb TNT
5,000 lb TNT
50,000 lb TNT
Pressure (psi)
1,000
100
10
1
0.5
0
100
200
300
400 500 600
Standoff (ft)
700
800
900 1,000
Figure 12.2 Blast Pressure Loading versus Standoff
between defended standoff and protection offered by conventional construction
is shown in Figure 12.3 (Smith et al. 2005).
Figure 12.3 illustrates the level of protection offered by a wide range of conventionally constructed buildings for a given setback. The left portions of the
bars indicate that no significant protection from blast effects is readily attainable
at these distances with conventional construction. The next area is an indication of a low level of protection. At these distances, conventionally constructed
buildings will typically sustain moderate to heavy damage. Occupants in exposed structures may suffer temporary hearing loss and injury from the force
of the blast wave and building debris fragmentation. Other assets may receive
damage from these effects. The third area is an indication of a medium level
of protection. At these distances, conventionally constructed buildings will generally sustain light to moderate damage. Occupants of exposed structures may
suffer minor injuries from secondary effects such as building debris. The rightmost area indicates a high level of protection. At these distances, conventionally
constructed buildings will generally sustain only minor damage.
12.2.1 Balancing Hardening with Standoff
A common question that designers face is “Where do I put my barriers?” Typical security personnel will declare that barriers must be as far away from the
protected asset as possible. Typical architects will respond that barriers intrude
310
DEFENDED PERIMETER
Medium-Level
Protection
No Protection
Low-Level
Protection
High-Level
Protection
Explosive Weight (lbs.)
1000
500
220
50
0
200
400
600
Standoff Distance (Feet)
Figure 12.3 Effects of Standoff Distance for a Typical Conventionally Constructed
Facility (explosive weight is in lbs of TNT equivalent)
on the appearance and function of the project and must be minimized. Typical
owners will say they want to be safe but, at the same time, not inconvenienced.
Typical community planning boards will want free, unrestrained access to the
site for community use. In most practical cases, especially for government facilities, there may be design criteria that prescribe a specific minimum standoff for
the defined threats. In other cases, there may be no such clear requirement. In the
end, politics or other nontechnical factors may also play a deciding role in the
selection and placement of barrier systems.
So what’s a designer to do? It is important to consider all aspects of the
design in determining where barriers are placed, the type of barriers to be used,
and the protection level that such barriers will provide. The site features may
play a role in determining the placement of the protected asset or building
within a protected site. From a protection engineer’s perspective, balancing
setback requirements with structural hardening and hazard-mitigating features
should be of primary concern. Designing to resist explosive effects increases the
STANDOFF
311
requirements of structural components; thicker walls, additional reinforcement,
and blast-resistant glazing and frames may be needed. An increase in explosive
weight or a decrease in standoff generally increases structural requirements. For
an efficient design, one must balance the effect of available standoff with the
effect of incorporating blast-resistant design and/or hazard-mitigation measures.
The types of perimeter protection should also be balanced with the types of
attacks considered in the risk assessment. A crash-rated barrier line is not needed
if the attackers are unwilling to risk drawing attention to themselves. However,
heavy barriers may be the only option to stop a truly determined attacker. Potential attacks can generally be reduced to three types:
r A true covert attack in which the attacker desires to completely escape notice and repercussion. Attackers of this type will avoid anything that draws
attention to themselves, such as high speeds and illegal parking. Eliminating public parking near the building and presenting a notable guard presence
will deter this type of attacker and provide standoff.
r An attack in which the aggressor is dedicated to the attack, but intends to
escape. This attacker will break traffic laws, but wants to get out of the
area before a police response can be mounted. Designated parking or even
illegal parking will not deter this attacker since he or she can flee the area
before a response is mounted. However, heavy-rated barrier systems are not
necessarily required, since this type of attacker will be deterred by the risk
of a potentially fatal crash or a significant delay in escaping.
r A suicide attack by an attacker who is willing to be captured or killed in
the attempt. If vehicle barriers are not in place, this attacker will drive onto
the site and through the front door of the building. Stopping this threat will
require significant and consistent hardening of the perimeter against vehicular attacks. Developing appropriate limits on achievable barrier penetration
is crucial. Sally ports and other redundant barriers may be necessary at all
vehicular entrances.
12.2.2 Balancing Costs
The effects of standoff provided by barrier systems on various structural and
nonstructural components are conceptually illustrated in Figure 12.4 (Smith
et al. 2005). This figure generally illustrates, at no specific scale, the typical
relationships between standoff and the cost of protection. A number of the
various components of incremental security cost are shown, including structural
and nonstructural component contributors. The relative magnitude and scale of
these relationships will vary from project to project.
For example, the cost associated with hardening the mailroom, loading dock,
and lobby to meet the design requirements may not vary with the available standoff to a vehicle-delivered bomb along the site perimeter. The cost associated with
progressive collapse considerations may also be constant with standoff, since it
312
DEFENDED PERIMETER
Incremental Cost of Protection $
is often treated as threat-independent by security design criteria. However, there
may be a point at shorter standoffs where the structural framing design is further impacted by the blast loading on the frame (above the progressive collapse
requirements), resulting in larger framing members and additional cost. Some
criteria may place limits on the design blast pressure and impulse for certain
secondary building components such as doors and windows, resulting in the relatively flat curves seen in the figure. However, some criteria do not use such restrictions and are likely to experience a significant cost increase at low standoffs.
The sum of the varying costs of hardening for the various components results in
the “cost of hardening” curve indicated on Figure 12.4.
One cost component that increases with increasing standoff is that for site
area and perimeter protection. For example, to provide increased standoff, the
distance to the defended perimeter must increase, thereby increasing the area of
Total Protection Cost
(hardening + land + perimeter)
Not To Scale
cost of land +
perimeter protection
cost of hardening
frame
progressive
collapse
windows & walls
other, mailroom, loading dock, lobby
20
50
limit
STANDOFF (ft)
RISK
High to
Catastrophic
Moderate
Moderate to Low
High to Moderate
Figure 12.4 Impact of Standoff Distance on Total Protection Costs (Compliments of
Applied Research Associates, Inc.)
STANDOFF
313
the site and the length of the perimeter that must be protected. Adding the cost
of hardening and the cost of land and perimeter protection results in the curve
indicated as “Total Protection Cost.” In the particular example shown in Figure
12.4, a standoff distance in the range of 30 to 50 feet would provide the desired
protection at minimum total cost. A project-specific analysis that examines these
variables can help in determining the optimal standoff for the proposed site.
12.2.3 Site Planning
Good site planning is the first step in ensuring that the desired standoff distances
are achieved at a given site. An effective site plan will minimize or eliminate
places where explosive devices can be concealed, will ensure that all activity
around the facility is subject to casual observation by the occupants or actively
monitored by a security force, and will discourage or prevent pedestrian or vehicular traffic in vulnerable areas.
There are an almost endless number of possibilities when it comes to implementing security features on a new site, and there is a great deal of room for
creativity. As shown in Figure 12.5, numerous types of barriers can be integrated
into the site plan. The implementation of landscaping and architectural features
can also have an important effect on the vulnerability of a facility. The positioning of roadways, parking, and sidewalks relative to the building can discourage
or encourage traffic. These features can also affect the likelihood of detecting
an attack or the time between detecting an attack and responding to it. In the
case of vehicular attack, roadway planning and landscaping features can limit
Figure 12.5 Integrated Site Plan
314
DEFENDED PERIMETER
the impact speed and approach angle of an attacking vehicle, resulting in lighter
barrier requirements. However, it is important that security be considered early
in the design process, when civil, architectural, and structural changes can be
made with minimal impact on ongoing designs in other nonsecurity disciplines.
Controlling the maximum impact velocity of an attacking vehicle is one very
important part of site planning, since the velocity term is squared in the relationship for kinetic energy and has a very large effect on the size and cost of
vehicular barriers. For example, doubling the impact speed increases the impact
energy for a given vehicle impact by a factor of four, thus greatly increasing the
design requirements and potential cost of a barrier system.
Using basic engineering and physics, there are formulas and methods available to estimate the maximum attainable impact speeds that a vehicle may obtain.
Performing such a study early in the planning process can significantly reduce
the costs associated with vehicle barriers.
Attainable Vehicle Speed on a Straight, Level Path The most basic vehicle
speed calculation is based on a straight and level path between the starting point
and the vehicle barrier. The final speed for such an approach can be calculated
using the following basic equation (Unified Facilities Criteria Program 2005b):
v2f = v20 + 2as
where: vf = final vehicle velocity
v0 = initial vehicle velocity
a = constant rate of acceleration
s = distance traveled
Initial vehicle velocity and distance traveled are easy concepts to grasp and
develop input values for. However, rate of acceleration can be difficult to determine. First, acceleration is very dependent on the weight and power output of
the vehicle. Average accelerations between zero and 60 mph can range from 25
feet per second squared for a modern sports car to 5.8 ft per second squared for a
2.5-ton commercial truck. Second, acceleration is not constant. Neglecting drag
and rolling friction, a vehicle producing a constant 200 hp at the rear wheels and
weighing 4000 lb will accelerate from 20 mph to 30 mph in 0.61 sec (average
acceleration = 24 ft/secˆ2). However, the same vehicle will require 1.34 seconds
to accelerate from 50 mph to 60 mph (average acceleration = 11 ft/secˆ2). As
can be seen, a 30-mph increase in speed reduced the rate of acceleration by
one half.
These numbers were obtained by solving the following equations through
time:
vi = v(i -1) + a(i -1) (dt)
ai = P/(mvi )
STANDOFF
where:
315
P = power
m = mass
vi = current velocity
v(i-1) = velocity at last time increment
ai = current acceleration
a(i-1) = acceleration at last time increment
dt = time increment
Slopes A downhill slope will decrease the required acceleration distance, and
an uphill slope will increase it. A correction factor can be obtained using the
following equation:
s /s = 1/[1 + (g/a) sin(theta)]
where:
s = acceleration distance needed to attain final speed on a sloped path
s = acceleration distance needed to attain final speed on a horizontal
path
a = acceleration
g = gravitational constant = 32.2 feet per second squared
theta = angle of slope (0 is level, positive is downhill slope)
(Unified Facilities Criteria Program 2005b)
Maintainable Vehicle Speed on a Curved Path Centrifugal force will result in
loss of control of a vehicle driving on a curved path if it exceeds the available
friction force on the tires. By equating centripetal force to friction, the following
equation can be derived for a vehicle traveling on a flat surface:
vs = sqrt( fgr)
where: vs = skid velocity
f = friction coefficient
g = gravitational constant = 32.2 feet per second squared
r = radius of curvature
(Unified Facilities Criteria Program 2005b)
The value of the friction coefficient, f , is highly variable. It depends on the
tire and its condition, the material and condition of the drive path, and any other
factors such as oil or water on the drive surface. The value of f should generally
fall between 0 and 1. For common vehicles under normal conditions, f = 0.6
is usually used. Values greater than 1 are possible for a dedicated race vehicle
on specially prepared tracks, but a value of f = 1 is generally safe when road
conditions are unknown, or a more conservative value is desired.
Curve Banking Banking can increase the achievable speed of a vehicle in a
curve and is commonly used in roadway design. The following equation adds an
316
DEFENDED PERIMETER
additional term to the curved path equation for calculating maintainable vehicle
speed on a banked curve, to cover that possibility (Hickerson 1967):
vs = sqrt[( f + tan(alpha))gr ]
where:
vs = skid velocity
f = friction coefficient
g = gravitational constant = 32.2 feet per second squared
r = radius of curvature
alpha = angle of embankment, positive for outer edge of curve higher
than inner edge
12.3 VEHICLE CONTROL BARRIERS
Since larger explosive weapons are difficult to carry and conceal on an individual, vehicles play a critical role in many explosive attacks, and vehicular control
should be considered when protecting a facility from large explosive threats. A
successful vehicle control system will prevent an unscreened vehicle from reaching a point where it can be used to damage or destroy the facility.
A well-planned site with good detection equipment and a properly trained
guard force can be effective in keeping what is generally referred to as a stationary vehicle-delivered threat away from a facility. That is, a vehicle driven
to the target, parked, and left by an aggressor intent on escape. However, barriers designed to defeat moving vehicular threats will be required to stop a more
determined aggressor willing to be captured or killed in a successful attack.
Barriers designed to defeat a moving vehicular threat (that is a threat actively
driven into the site or facility) must be designed to arrest the energy of the attacking vehicle. For most standards, kinetic energy is used as a basis for establishing
vehicle barrier performance requirements. The gross mass of a vehicle (vehicle
mass plus the mass of explosives or any other cargo) and its maximum attainable
speed at the point of impact define the kinetic energy that must be absorbed by
the barrier. Kinetic energy can be expressed as:
K E = 0.5 mv2
where: KE = kinetic energy
m = gross mass of vehicle
v = velocity of vehicle
12.3.1 Crash Testing
Both the U.S. Department of State (DoS) and the U.S. Department of Defense
(DoD) rate barriers based on full-scale crash tests conducted by independent test
laboratories or government-approved facilities. The testing standard traditionally used for these tests is designated SD-STD-02.1, Specification for Vehicle
VEHICLE CONTROL BARRIERS
317
Crash Test of Perimeter Barriers and Gates (U.S. Department of State 1999).
The DoS test standard rates the anti-ram barriers within three categories: K4,
K8, and K12. The K rating corresponds to the kinetic energy of the impacting
vehicle at the moment of impact. A K12 rating corresponds to impact energy of
1,200,000 ft-lb. These tests are conducted using a flatbed truck with a designated
mass of 15,000 lb. The standard originally provided an additional L rating that
designated the level of penetration achieved by the vehicle. However, in March
2003, with the release of Revision A, acceptable penetration was limited to 1 m,
eliminating the need for an L rating. All previous barriers where the penetration
exceeded 1 m were designated unacceptable by the revised standard, and the L
rating was dropped from use. While significant penetration is not allowed by
the DoS standards, penetration may be permitted in other applications especially
where large setback is available.
ASTM has also developed a test standard, with input from both the U.S. Department of State and the U.S. Department of Defense, that is beginning to grow
in acceptance in the barrier industry. This standard, Standard Test Method for
Vehicle Crash Testing of Perimeter Barriers (ASTM F2656-07) is likely to grow
in use and is gradually being accepted by both the U.S. Department of State and
the U.S. Department of Defense.
12.3.2 Crash Modeling
There is no one-to-one correlation between the kinetic energy and a static design
load. The impact forces vary significantly through time and are a function of the
crushing of the vehicle as well as the strength and stiffness of the barrier. One
method for calculating barrier response is to equate the energy in the vehicle
(KE) with the work done on the barrier and the vehicle:
KE = Wb + Wv
Wb = F(x) · d xb
Wv = F(x) · d xv
where: KE = kinetic energy
Wb = work done on the barrier
Wv = work done on the vehicle
F = force generated between the barrier and the vehicle
xb = displacement of the barrier
xv = displacement of the vehicle
Calculating work on the barrier and the vehicle depends on obtaining good
force and displacement histories. This is difficult to accomplish with simple
models, given the complex interactions that occur between the barrier and the
deforming vehicle during impact. A conservative simplified analysis can be performed using a single-degree-of-freedom (SDOF) model of the barrier, if the
work done on the vehicle is set to zero. However, an SDOF design will generally
318
DEFENDED PERIMETER
Figure 12.6 Comparison of Vehicle Impact Test to Finite Element Analysis (Compliments of Applied Research Associates, Inc.)
be very conservative. Testing has shown that the real force and displacement histories that occur during impact are quite complex, and the energy absorbed by
the vehicle can be significant.
Given the need to make assumptions that ensure conservatism, barrier system designs resulting from simplified analysis can be significantly overdesigned.
Therefore, barrier systems are generally certified through full-scale crash testing
or sophisticated analysis.
Properly performed nonlinear finite element analysis (FEA) techniques have
shown good correlation to test data and can be an effective means of developing
new barrier systems. FEA models can replicate many of the complex interactions
that control the force and displacement histories in actual crashes. Examples of
an actual vehicle impact compared to a successful pretest finite element analysis
are shown in Figures 12.6 and 12.7.
Finite element analysis is a powerful tool, but it should generally be supplemented with crash testing to verify the analysis results. When such testing is not
possible, a qualified firm or consultant who is independent of the manufacturer
should be contracted to perform or review such analysis. FEA is most effective
Figure 12.7 Front View of Impact versus Analysis Result (Compliments of Applied
Research Associates, Inc.)
VEHICLE CONTROL BARRIERS
319
Figure 12.8 Low Wall (Compliments of Applied Research Associates, Inc.)
when the analyst has access to crash test data for similar systems and vehicles in
order to validate the FEA model performance. Therefore, the more the proposed
system differs from known test parameters, the more doubt and scrutiny should
be given to the results. Ideally, the analyst will be able to provide FEA results
that demonstrably replicate tests for similar barrier systems.
12.3.3 Walls
Walls and retaining walls, in particular, are some of the most robust crash barrier systems available. Commercial wall systems are available that have been
designed and tested to stop a 60,000-lb vehicle traveling at 50 mph (KE =
5,000,000 ft-lb). Reinforced concrete is a common construction material for
crash-rated walls. The wall thickness is generally between 12 and 24 inches.
Crash-rated walls have traditionally risen 36 inches above grade, but recent testing and analysis have supported the use of 30-inch-tall barriers for stopping typical cars and some trucks. Figure 12.8 illustrates what can happen when the wall
height is set too low.
12.3.4 Bollards
Bollards are one of the most common vehicle barrier systems in use. Bollards
are usually cylindrical posts spaced close enough to prevent vehicle passage and
are typically constructed of steel pipe anchored in a reinforced concrete foundation. A few bollard systems are designed with foundations less than 1 foot deep,
but most have deeper foundations. Four feet is typical. Bollard spacing is also
frequently set at 4 ft, but bollards can in many cases be effective at spacing up
to 5 ft. Spacings beyond 5 ft are generally avoided, since small cars or trucks are
capable of squeezing between bollards spaced farther than 5 ft apart. A typical
passive bollard design is shown in Figure 12.9.
The bollard shown is capable of a K4 rating. K12-rated bollard designs are
also available.
320
DEFENDED PERIMETER
Figure 12.9 Eight-Inch-Diameter Fixed Bollard System
Active bollards can also be obtained. The most basic can be manually detached from the foundation and carried away (Figure 12.10).
The more advanced systems can be retracted into the foundation, using hydraulic or electric motors. Figure 12.11 illustrates one such retractable hydraulically operated bollard system.
12.3.5 Active Wedge
Wedges are a common type of active barrier and are capable of achieving K12
ratings. Wedges are typically installed to be flush with the road surface when
lowered. When the barrier is raised to stop traffic, steel plates pivot up from the
road surface to form a wedge shape. Wedges are typically operated electrically
or hydraulically. Many variations are available. Figure 12.12 illustrates one common configuration of wedge barrier.
12.3.6 Beam Barriers
Beam barriers generally swing, pivot, or slide out of the path of the vehicle to
permit traffic flow. Beam barriers generally do not achieve impact ratings as high
as those of other active wedge barriers, but they generally cost significantly less
and are easier to operate. The photo in Figure 12.13 shows a beam barrier.
VEHICLE CONTROL BARRIERS
Figure 12.10 Removable Bollard
Figure 12.11 Operable (Pop-Up) Bollards
321
322
DEFENDED PERIMETER
Figure 12.12 Wedge Barrier
Figure 12.13 Beam Barrier
VEHICLE CONTROL BARRIERS
323
12.3.7 Cable-Based Systems
Cable-based systems generally consist of horizontally spanning steel cables anchored to the ground through anchor posts or other means and positioned to
capture an attacking vehicle. When a cable system is impacted by a vehicle, the
intermediate anchor posts respond progressively away from the point of impact
until the vehicle is arrested. Cable-based barrier systems generally allow more
penetration than other more rigid barrier systems. However, they also generally
require lighter or more widely spaced foundations. Another advantage of cablebased systems is their ability to absorb the energy of the impact in a fashion that
limits vehicle damage and injury to the vehicle occupants. The anchored cables
can also be integrated into various fence systems to prevent intrusion of people
as well as vehicles. See Figure 12.14.
Cable barriers are a good choice for large perimeters where a few feet of
penetration is acceptable in exchange for reduced installation costs. Cable-based
systems are most effective for protecting long stretches of perimeter, since they
Figure 12.14 Cable Fence Barrier
324
DEFENDED PERIMETER
Figure 12.15 Concrete Planter
perform best when splice points are widely spaced. Fences with cables are
common at airports throughout the United States. These systems are generally
vulnerable to impacts near corners and ends, where the impact forces cannot be
transferred to adjacent anchor posts through tensile-membrane action. Special
detailing or supplemental barriers may be required near corners to ensure a
consistent level of protection.
Active cable-based systems also exist and typically consist of a cable net that
spans the road to prevent traffic flow. The cables can be relaxed to allow vehicles
to pass over the cable net or tensioned to activate the barrier.
12.3.8 Planter and Surface Barriers
Surface barriers typically consist of large containers filled with soil, water, sand,
or some other readily available material. One of the more common and aesthetically pleasing barriers in this category is concrete planters (see Figure 12.15).
True surface barriers (ones that do not include a foundation) initially absorb
the energy of impact through momentum transfer and must have significant
mass to perform well. Lighter surface barrier systems may allow significant
penetration before stopping a vehicle. Figure 12.16 shows the aftermath of a
real-world collision with a relatively lightweight planter system. In this case, the
planter was driven nearly to the building before coming to a full stop. Ground or
concrete anchors are available for such systems and will significantly reduce the
penetration distance.
12.3.9 Berms, Ditches, and Other Landscaping Features
Berms, ditches, and hillside cuts, as well as trees, boulders, and other large
objects can be effectively used to stop vehicles from penetrating a restricted
PEDESTRIAN CONTROL BARRIERS
325
Figure 12.16 Concrete Planter Penetration
boundary. The authors are not aware of detailed analysis or testing that verifies
the kinetic energy resistance of this category of barrier, but these systems can
provide significant vehicular control. Figure 12.17 diagrams several commonly
used barriers in this category.
Triangular ditches, trapezoidal ditches, and hillside cuts are easy to construct
and are effective against a wide range of vehicle types. The designs presented in
Figure 12.18 are based on typical vehicles with respect to tire size and ground
clearance. Dedicated off-road vehicles may defeat these systems, and careful
consideration should be given to likely threats. It is also important to consider
the speed and trajectory of an approaching vehicle when implementing any of
these systems. For example, it may be possible for a vehicle to jump a ditch
or hillside cut, given an ideal approach. The width of the ditch may need to be
increased where a high-speed approach is possible.
Boulders, trees, and other vegetation can also be used to stop vehicles. One
such application is shown in Figure 12.18.
12.4 PEDESTRIAN CONTROL BARRIERS
Pedestrian control is designed to direct and moderate the flow of people around
or through a facility. Good pedestrian control will deter or prevent visitors or
uncleared personnel from entering areas of the facility vulnerable to attack. It is
important to remember that a properly placed man-portable explosive device in
close contact with a protected asset or structure can have a devastating effect.
Good site planning is the first step in pedestrian control and may be the
only pedestrian control system implemented outside the building envelope for
many facilities. An effective site layout will eliminate places where explosive
326
DEFENDED PERIMETER
Figure 12.17 Excavation Barriers
BLAST WALLS AND BERMS
327
Figure 12.18 Trees and Boulders Used as Barriers
devices can be concealed, will ensure that all activity around the facility is
subject to casual observation by the occupants, and will discourage pedestrian
traffic in vulnerable areas without drawing attention to those areas. Sidewalks,
landscaping, lighting, and signage can all contribute to achieving these goals.
True physical pedestrian barriers are prudent for facilities where there is a
significant risk of pedestrian attacks. Fencing or high walls are the most common
type of pedestrian control barrier for the building perimeter. Fencing and walls
come in many sizes and architectural styles, and each style carries a different
delay factor based on height and structure. However, for the purpose of security,
fencing or walls should generally be a minimum of eight (8) feet high and be
durable enough to impede or withstand attacks from at least simple hand tools
such as bolt cutters. In some cases, fences and walls can be built to also act
as vehicle barriers. An example of this is shown in Figure 12.19. This fence
application includes heavy cables secured to massive, buried anchors that serve
to effectively stop a vehicle-borne threat.
12.5 BLAST WALLS AND BERMS
Obtaining adequate standoff to reduce the blast loads to an acceptable level is not
always possible. One way to reduce loading is the implementation of blast walls.
Properly implemented blast walls and revetments can mitigate the peak air-blast
pressure and impulse from an explosion.
When placed across the path of the air-blast shock wave, the barrier may
reflect a portion of the incident pressure back toward the explosion, or cause
the blast to diffract over the barrier. The diffracted pressure will be significantly
328
DEFENDED PERIMETER
Figure 12.19 Anti-Climb Fence with Integrated Catch Cables
reduced for some distance behind the barrier before the shock reforms to its
original intensity. The area of effectiveness behind the barrier is controlled by
the dimensions of the blast wall or revetment, the size of the explosive device,
and the relative locations of the bomb, the wall, and the building.
Properly designed blast walls can also capture vehicle and bomb fragments
before they reach the building. This can be an important consideration, especially when dealing with military-grade weapons designed to generate hazardous
fragments.
Proper hardening of the wall is an important consideration in implementing a
blast wall. Testing has shown that walls located close to the explosive device can
break up and become dangerous fragments during an explosion. In many cases,
the energy imparted to the facility by wall debris can negate the benefits provided
by the reduction in air-blast loads. Using wall types that are inherently resistant to
generating hazardous fragments is an option. Lightweight containers filled with
water and sand are examples of systems with low secondary fragmentation risks.
The phenomena that occur behind a blast wall and that generate load reductions are very complex. A site-specific analysis is required to properly locate
the wall and calculate the benefits. An improperly implemented blast wall will
provide little benefit, and it is possible for a blast wall to actually increase the
blast loads on a structure. A documented procedure for analyzing the benefits of
blast walls is presented in UFC 4-020-03, Security Engineering, Final Design
(Unified Facilities Criteria Program 2005a).
The primary sources of information about the benefits of blast walls and
berms are experiments conducted on blast walls and a study on revetments.
These studies only covered a limited range of devices and barrier heights, but
the following trends have been observed. The wall must be close to the building
REFERENCES
329
or the explosive charge to achieve useful load reductions. For walls placed
close to the charge, peak pressures are reduced dramatically directly behind
the wall but gradually approach the free-field blast at large distances. There are
some combinations of large distances between the bomb and the wall and short
distances between the wall and the building that may cause an increase in the
reflected impulse due to multiple reflections of the blast wave between the wall
and the structure.
The shape of the wall or barrier and its height have a significant effect on the
load reduction. Walls and revetments with faces oriented perpendicular to the
shock wave (and the ground surface) should be a goal, since this arrangement
provides by far the most benefit. Mounded or sloped barricades (i.e., typical berm
construction) provide little reduction in blast loads.
The current blast wall data are based on situations where the wall height was
selected by taking the cube root of the charge weight in pounds and multiplying
by values between 0.8 ft and 1.3 ft. For example, the cube root of 500 lb is 7.9.
Therefore, 6-ft to 10.5-ft walls fall within the range of the data. Trends in the
data show that the wall is most effective if the bomb is within 1 wall height of
the wall and within approximately 10 wall heights of the building. Ranges from
the wall to the building greater than 20 wall heights generally result in negligible
reductions.
REFERENCES
Hickerson, Thomas F. 1967. Route Location and Design (5th ed.) New York: McGraw
Hill.
Smith, Joseph L. et al. 2005. Protective Design and Security Implementation Guidelines.
Prepared by Applied Research Associates, Inc. for the U.S. General Services Administration.
Unified Facilities Criteria Program. 2005a. Security Engineering, Final Design (UFC
4-020-03). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria
Program.
. 2005b. Selection and Application of Vehicle Barriers (UFC 4-022-02). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Department of State. 1999. Specification for Vehicle Crash Test of Perimeter Barriers and Gates (SD-STD-02.1). Washington, DC: U.S. Department of State.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
13
Blast-Resistant Design of
Building Systems
Scott Campbell and James Ruggieri
13.1 BACKGROUND
Public awareness of the need to protect structures against blast effects has risen
sharply since September 11, 2001. A number of public agencies are engaged
in research and development of blast-resistant technology, and its application to
domestic and foreign structures that are potential targets of terrorist activity. The
federal agency responsible for developing technology in this area is the Defense
Special Weapons Agency/Technical Support Working Group (DSWA/TSWG),
which receives funding from Congress through Department of Defense (DoD)
appropriations. This funding is used to support physical testing of candidate
blast-resistant design concepts, computational modeling, and understanding
of phenomenology. A range of government agencies and private consultants
participates in the program, whose results are available to the General Services
Administration (GSA), the DOD, and other U.S. and foreign government
agencies. The technology derived from this and similar previous programs is
currently being implemented for the protection of U.S. Secret Service headquarters and FBI laboratory in the Washington, D.C.. area, the federal courthouse
in Brooklyn, NY, and the passenger terminal at Midway Airport, Chicago, to
name only a few locations. The U.S. State Department plans to request between
$1 and $3 billion to raise security at U.S. embassies to the standards set by the
Department in the 1980s.
The events of 9/11, as well as similar follow-on events, have introduced
new performance requirements for all building systems. New performance
requirements for such structures have introduced a need to revisit and modify
the traditional assumptions governing building design. New systems, including
atmosphere segregation systems (i.e., positive pressure, collective protection,
and hazardous agent monitoring systems, etc.) as well as changes to the
environment, including consideration to blast and shock, introduce constraints
not previously considered in the design development of building mission and
life-safety systems.
331
332
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Although existing commercial consensus standards, such as NFPA 70A (National Fire Protection Association 2005), have historically performed well for
general “nontarget” buildings and facilities, they do not address the special considerations now facing “target” buildings and facilities serving critical infrastructure needs. Therefore, new practical and readily expedient methods need to be
identified that could better fulfill this new environment.
13.2 INTRODUCTION
The detonation of explosives near or in a building affects not only the building
facade and structure, but also the building contents. In seismic events, the
improvement in earthquake-resistant structural design has led to a significant
decrease in structural damage. However, the result is that nonstructural damage
now accounts for a majority of the costs associated with earthquakes. Similarly,
increasing the blast resistance of building structures will reduce the associated
damage, but may increase the relative importance of nonstructural component
damage. Significantly, a building may be structurally undamaged, but rendered
useless due to damage to HVAC, electrical, plumbing, and other systems.
Therefore, a balanced design approach is suggested wherein the building
systems are designed to withstand either the appropriate blast loads or resulting
structural movements.
Very little information is available regarding the design of building systems to
resist the effects of explosions. This lack of design guidance has resulted in the
use of better-developed seismic guidelines, such as the Uniform Building Code
(UBC) (International Code Council 1997) or International Building Code (IBC)
(International Code Council 2006), the use of extremely conservative designs,
or the neglect of the protection of equipment and systems. Traditionally, equipment has been designed for blast by being placed in shelters or internal to the
building and checking acceleration resistance against expected levels. This remains the best practice, although particularly rugged equipment—for example
pumps—may be left exposed with anchorage designed for the applied loads.
Third-party testing by nationally accredited conformity assessment organizations, such as Underwriters Laboratories (UL), is often used to verify certain performance characteristics of a system element. Traditional testing often includes
verification of desired material properties such as strength, ductility, chemistry,
and composition; and random sampling procedures are used to audit performance of the subject components for use, in comparison to design requirements
and/or regulation. Examples of this include equipment operability, controllability and speed, current-carrying capacity, insulation resistance, power required,
power developed, vibration, and many others. In most cases, the requirements invoked for electrical equipment and devices used in building systems only address
electrical shock safety (e.g., electrocution risk) and risk of fire. However, such
criteria presume a benign environment, not subject to risk of blast. The standards
DESIGN CONSIDERATIONS
333
and building code organizations readily admit that such documents only identify
a consensus for a common denominator that speaks to the minimum acceptable
performance or criteria, assuming a safe and relatively comfortable environment.
Such documents do not address blast. Although there is one U.S. national standards development activity presently underway to provide enhanced criteria to
harden building electrical systems to the effects of blast—ASCE/AEI Recommended Hardening Techniques for Control, Communication and Power (C2P)
Systems of Critical Facilities—there is nothing available in the nonmilitary community at this time that can be used for design guidance.
Explosions generate loads on building systems through three basic mechanisms. Direct blast pressures can impact systems exterior to the building from
explosions outside the building envelope and from internal explosions. Indirect
pressures are generated when a blast external to the space where the equipment is located leaks into the room through openings. Finally, ground shock and
air shock generate loads on building systems through motions developed in the
structure. All three types of loads are considered in this chapter. Note also that
fragments can pose a severe threat to building systems and equipment. However,
calculation of fragmentation and the resulting damage to nonstructural components is beyond the scope of this section (see Chapter 8).
The general design approach for building systems is to determine the load
levels, both blast pressures and base accelerations, for the chosen location. The
equipment and anchorage capacities are then checked against the design load.
If the equipment’s allowable load is exceeded, then either it must be moved or
protected. Each of these steps is detailed in subsequent sections of this chapter.
13.3 DESIGN CONSIDERATIONS
The key philosophy for providing continuity of services for mechanical and electrical systems in buildings is SHR: Separation, Hardening, and Redundancy.
Segregation of critical and vital circuits from normal services provides design
economy, since hardening-type construction methods can be focused on only
those specialized areas. Judicious use of steel conduits, pipes, and ducts associated with life-safety systems provides increased likelihood of survivability for
these systems, and a greater likelihood that emergency systems will remain operational post-event to assist rescuers in the evacuation of the building, and/or
provide continuity of critical business services.
Although each design situation is unique, it is possible to provide some general guidance regarding blast-resistant design of equipment and systems. This
section will discuss choosing appropriate design goals, specific points of concern for the different loads that can be imposed, and some guidelines regarding
select types of equipment and systems. However, good engineering practice and
judgment are the best approach for these situations, given the amount of uncertainty present in this field.
334
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
13.3.1 Level of Protection
A critical step in the design of building systems for blast load is to determine
the goal of hardening. In general, there are two levels of performance normally
considered in system blast-resistant design:
1. Life Safety: Also called position retention. This design level is intended to
keep the equipment in place, with no large pieces or components coming
loose and potentially injuring personnel. The equipment or system is not
expected to remain operational after the event.
2. Continued Operation: As the name implies, the equipment is expected to
remain operational after the event. This might entail resetting the equipment, but no major repairs should be required to maintain basic operational
capabilities.
Continued operation of equipment subjected to blast or shock loads is often
difficult to predict. Seemingly minor failures can lead to a shutdown and loss
of critical systems. In general, only inherently rugged equipment can be considered for continued operation if directly exposed to blast pressures. Less rugged
equipment may be designed to remain operational for reduced pressures due to
protection by shelters or interior location. If the continued operation of a system
is required, it is essential that all aspects of the loading condition be considered,
and testing of the equipment is likely to be necessary.
The life-safety level of design is easier to achieve, as no evaluation of the
working parts of the equipment is required. It is only necessary to analyze the
equipment structure and anchorage. Note, however, that the analysis includes all
external components, including the sheet metal shell of the equipment, if any, to
ensure that they remain attached to the equipment. Large internal components
that have the ability to break out of the shell must also be considered. Another
response quantity that must be considered for suspended systems is deflections.
A system may be capable of resisting the pressure or acceleration load on its
own, but large deflections may cause an impact with other equipment and lead
to an anchorage or equipment failure.
13.3.2 Blast Pressures
The best way to protect equipment and systems from the effects of blast is to
locate them in areas protected from direct air blast. If the building envelope is
properly designed to resist blast effects, location of equipment in mechanical
rooms without openings to the exterior of the building will preclude pressure
loading from external events. However, equipment that is located inside protective structures will often be subjected to blast pressure that leaks in through openings in the walls and/or roof. Likewise, properly securing hardened mechanical
rooms, and thus limiting access to authorized personnel, will significantly reduce
the likelihood of internal blast loads on the equipment. However, concentrating
DESIGN CONSIDERATIONS
335
mechanical equipment in a single location will render the overall system more
vulnerable to unauthorized access if proper precautions are not maintained.
Removal of equipment from vulnerable areas is not always possible, and is
likely not feasible for all distributed systems. In these situations, it is necessary to
require design to higher than desired levels of pressure loading. This may entail
providing barrier walls, if not complete shelters, or testing equipment to the anticipated design load. Particularly critical distributed systems may be enclosed in
hardened runways, or redundant systems located elsewhere in the building may
be provided. Obviously, these additional precautions can be extremely expensive, and care should be taken early in the design process to eliminate or reduce
to a minimum the number and types of systems that require special design.
13.3.3 Shock Induced by the Structure
The impact of blast-generated air and ground shock on the structure will produce
motions that are potentially damaging to the building systems. The resulting
structural accelerations can be estimated using the methods outlined later in
this chapter. After the structural motions are calculated, the resistance of the
equipment to shock loads can be checked. Typical values of equipment shock
resistance are given in Table 13.1. One major caveat applies to the use of these
values. In many cases they are based on testing that is decades old. In the interim
many changes have occurred in the design of some equipment. The tabulated
values may no longer be applicable, and caution should be used in applying
them to modern equipment. Significant testing is often required for equipment
used in the military and the nuclear industry, and manufacturers may be able
to provide capacities for specific pieces of equipment. The nuclear industry
Seismic Qualification Utility Group (SQUG) has a large database of equipment
response during earthquakes, but these data are not generally available to
the public.
A common practice is to specify design loads for the equipment in terms of an
equivalent earthquake load. This is particularly true where it is desirable that the
actual load source remain confidential. It is perfectly acceptable to give only an
equivalent “g” acceleration value. The acceleration level to be specified can be
determined using the methods detailed in Chapter 7 or a more detailed calculation that considers the stiffness and mass distribution in the structure. However,
there are several caveats that must be observed.
First, unless a dynamic analysis of the building is performed and the floor
accelerations specified as time histories, the frequency content of the input motion to the equipment is lost. Second, it is not correct to specify a design to,
for example, 0.75g using the 2006 International Building Code (IBC). The IBC
method is used to determine earthquake loads for anchorage design and equipment qualification. A base ground acceleration is used that is then modified to
account for location within the building, natural frequency of the equipment,
ability of the equipment to undergo nonlinear deformations without failure, and
the importance of the equipment.
336
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.1
2007)
Selected Equipment/System Fragility Level Guidelines (ASHRAE
Accelerations (g)
Equipment/System Type
No Damage Probable
Minor Damage Probable
Pumps up to 100 hp
4.5
8
Control Panels
3.5
8
Fans up to 100 hp
4
9
Welded steel piping (18 in. dia max, Sch
30 and up)
4
8
Air-handling units
4.5
10
Duct, steel
4.5
9
Motor control centers
2.5
N/A
Uninterruptible power supplies
3
N/A
Cable trays (up to 36 in.)
4.5
8
Chillers
Centrifugal
4
N/A
Absorption
4.5
N/A
Screw
3.5
N/A
Air-cooled
3.0
N/A
3.5
N/A
Boilers
There are several contributing factors as to why it is not good practice to
apply this methodology to in-structure shock loads. First, earthquake loads on
equipment depend not only on the size of the earthquake, but also on the vertical
location within the building. If the floor accelerations have been determined, and
that is what is being specified, it must be made clear that the factor for vertical
location (an increase in load of as much as three times) should be disregarded in
the calculation. It is also not clear if the other factors are applicable. However, the
most critical point is that many parts of the country are exempt from providing
seismic restraint to equipment regardless of the specified acceleration level. For
these reasons it is best to provide an acceleration level, either ground or floor
level; require use of the building code load equation as appropriate; and explicitly
state that no exemptions are permitted.
Shock Isolation: Isolation of equipment to reduce the effects of shock loading has been used successfully on many projects. Isolation systems work on
the principle that if the loading and isolation system’s natural frequencies vary
greatly, the isolated equipment will undergo motions much less severe than the
base structure. The amount the motion is reduced depends upon the ratio of the
forcing frequency to the natural frequency of the isolation system (frequency
ratio), as illustrated in Figure 13.1. Thus, to achieve a high level of reduction,
the isolation system should have a much lower natural frequency than the input
DESIGN CONSIDERATIONS
337
Figure 13.1 Transmissibility of Motion for Isolated Systems with 5% Damping
motion. However, this leads to a soft isolation system (low stiffness) and potentially high static deformations and problems with coupling between displacement
and rotation deformation modes. A balanced system using a properly selected
isolator, additional weight (inertia mass), and possibly supplemental damping
can help to alleviate the problems associated with soft isolation systems.
There are several types of isolators that are commonly used, including
steel springs, wire coils, air-springs, elastomeric mounts, and damping devices
(Figure 13.2). Most of the isolators can be used in either direct-mounted configurations, where the inertia mass and equipment are placed on top of the isolators,
or in a pendulum configuration where the load is hung from the isolators. In
either case, the systems are designed to carry the gravity load and accommodate
the deformations expected during a shock load. The equipment thus remains
relatively motionless while the supporting floor undergoes large motions, with
the isolation system deforming while maintaining vertical load capacity. Shock
isolation systems are available from a number of manufacturers who provide
design services along with the isolation systems. The project engineer must
determine the level of isolation required and the input motions, while the
manufacturer will choose the correct isolator to achieve the desired results.
13.3.4 Equipment/System Anchorage
The anchorage of equipment to the structure must be checked for all loading
cases and desired levels of performance. The procedure is relatively simple: Analytically apply the load to a mathematical model of the equipment, calculate
338
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
CMM Plate
Inertia Base
Kinetics
PSS Mount
(a)
(b)
Figure 13.2 Examples of Spring Isolators for Shock Isolation: (a) Spring isolator in
a pendulum configuration (Kinetics Noise Control, Inc.); (b) Wire rope isolators (VMC
Group, Inc.); (c) Air-spring isolator (Kinetics Noise Control, Inc.); (d) Elastomeric isolators (Kinetics Noise Control, Inc.)
the forces at the support location, and design the anchorage to accommodate the
maximum forces. The equipment is usually assumed to be rigid for the purpose
of calculating anchorage forces, since the dynamic effects are not considered.
If desired, a more detailed model that considers the mass and stiffness of the
equipment may be used. The load is applied at the center of the exposed surface for pressure loads, and at the center of mass for accelerations. It may also
be necessary to account for nonsymmetric loading and/or mass by varying the
load application point. In addition, the load should be considered to act in all
DESIGN CONSIDERATIONS
(c)
(d)
Figure 13.2 Continued...
339
340
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
possible directions. Note that this is not simply all directions, as in seismic design, since the explosive location may be restricted and therefore limit the load
direction. All seismic design guidelines should be observed regarding limitations
on the type of anchors. For example, power-actuated fasteners should not be used
unless specifically qualified, and undercut anchors are required for equipment
over 10 hp that is not vibration-isolated. In addition, the structural component to
which the equipment is mounted must also be capable of carrying the anchorage
force. Details on the anchorage calculations are provided in the design examples.
13.3.5 Placement of Critical Systems Equipment and Control Stations
Unsecured areas such as the lobby, loading dock, mailroom, garage, retail, and
other public access areas must be separated from the secured areas of the building, particularly areas of a building housing switchgear, computer server data
centers, centralized command centers, and machinery plant control and monitoring stations. Ideally, unsecured areas are placed exterior to the main building or
along the edges of the building. Occupied areas or areas providing emergency
or critical functions should not be placed immediately adjacent to the lobby, but
should be separated by a substantive buffer area such as a storage area or corridor.
From an operational security standpoint, it is important to restrict and control access to air-intake louvers, mechanical and electrical rooms, telecommunications
spaces, and rooftops by means of such measures as visitor screening, limited elevator stops, closed-circuit television (CCTV), detection, and card access-control
systems.
13.3.6 Staffing and Building Operations
How building staffing (i.e., managers, maintenance crew, boiler-men, security
personnel, attendants, etc.) is determined can make a substantial difference in
minimizing loss of life and minimizing injury, as well as improving the ability of a facility to sustain continuity of critical services in the event of a blast.
Determining staffing is integral to the design-development process of building
systems. For instance, automation and systems monitoring rely on certain assumptions regarding staffing. Increased use of centralized equipment/systems
health-determining methods, such as computerized condition-based monitoring
(vibration, temperature, differential pressures, etc.) significantly reduces the need
for periodic roving manned inspections.
An excellent model for effective staffing can be taken from the military. U.S.
warships are designed to absorb a certain number of strikes to minimize loss of
life, ensure continued reliable maneuvering and control (survival of steering gear,
propulsion systems, etc.), and ensure reliable operation of both defensive and
offensive systems. Such vessels rely on an architecture of self-sufficiency using
highly trained crews who are assigned damage control functions in addition to
their normal duties (e.g., cook and fireman). As with warships in battle, buildings
DESIGN CONSIDERATIONS
341
in the face of a civilian mass casualty incident need be self-sufficient, since it is
unlikely that first responders will be able to respond promptly, if at all.
Many elements of the military model can be easily adopted by the civilian
community, and much can be gained by investing in certain added training and
emergency duty assignments for critical personnel. Although using better-trained
personnel may increase operational costs, strategic optimization of variable resources will offset cost increases. For instance, since such personnel are better
trained in building systems, they could also perform certain maintenance and
repair functions routinely outsourced to specialists—serving not only to save
money but likely to reduce downtime of building systems.
The availability and familiarity of selected record drawings to such designated
personnel (i.e., electrical one-line, list of feeders and mains, equipment location
drawings, etc.) and the periodic drilling of personnel promote familiarity of systems while also identifying weaknesses in capability.
13.3.7 Construction of Hardened Spaces
Spaces dedicated to housing critical equipment should be constructed to withstand the effects of blast and fire in unsecured adjacent spaces. Historically, the
preferred material for explosion-mitigating construction has been cast-in-place
reinforced concrete. Reinforced concrete has a number of attributes that make it
the practical construction material of choice. The use of the new tough polymer
spray-on coatings on exposed concrete surfaces, as commonly used to protect the
R
R
beds of commercial pickup trucks (e.g., Line-X , Rhino .) checks the release
of airborne debris in the event of blast. Outfit and furnishings located in secure spaces should be selected to minimize fire load, while electrical equipment
should be mounted on resilient foundations to minimize the effects of shock.
Depending upon the function and equipment located in the secure space, consideration should be given to automated conflagration control systems, such as CO2
deluge systems, the use of fire-retardant cable and cable systems, and the use
of intumescent coatings. Positive pressure systems and/or gas-tight construction
should be considered in critical spaces to control air flow and thwart and control
undesired migration of contaminants.
13.3.8 HVAC and Plumbing Systems
Equipment that is deemed essential should be placed in secure locations. That
is obviously not possible for all system components, as they must be distributed
throughout the building. Because of that, design of HVAC/plumbing systems
usually focuses on position retention (adequate anchorage). Systems that service
areas that must remain functional after an event—for example a safe haven or
control center—should be served by redundant systems or have the equipment
and delivery systems placed in secure areas and hardened runs. The high cost
associated with these countermeasures necessitates a thorough evaluation of the
expected threat and the consequences of system failure.
342
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Air-intake locations should be located as high up in the building as practical,
to limit access and reduce the blast pressure level. Generally, air intake should
be located in secure areas. Systems that serve public access areas such as mailreceiving rooms, loading docks, lobbies, freight elevators/lobbies should be isolated and provided with dedicated air-handling systems capable of 100% exhaust
mode. Air-intake locations and fan rooms should be coordinated with the security
surveillance and alarm system so that surveillance is provided and unauthorized
access can be responded to by the security force.
Building HVAC systems are typically controlled by a building automation
system, which allows for quick response to shut down or selectively isolate
air-conditioning systems. These systems should be coordinated with the smokecontrol and fire-alarm systems to allow for automated, proper response to fires.
Access to building mechanical areas (e.g., mechanical rooms, roofs, elevator equipment access) should be restricted and monitored. These areas provide
access to equipment and systems (e.g., HVAC, elevator, building exhaust, and
communication and control) that could be used or manipulated to assist in smoke
evacuation after a blast or during a chemical, biological, or radiological (CBR)
attack. Additional protection may be provided by including these areas in those
monitored by electronic security and by eliminating elevator stops at the levels
that house this equipment. For rooftop mechanical equipment, ways of restricting (or at least monitoring) access to the roof that do not violate fire codes should
be considered.
Particular care should be taken to ensure that HVAC equipment and attached
ducts/pipes can accommodate any expected deformations. This generally means
using flexible connectors between duct/pipe and equipment, an example of which
is shown in Figure 13.3. In addition, distributed systems often cross areas with
Figure 13.3 Example of Flexible Connectors Near Equipment Attachment Points
(Courtesy of Kinetics Noise Control, Inc.)
DESIGN CONSIDERATIONS
343
large differential motions, including stories with different floor motions and
building joints. Attachments at these locations should also provide adequate deformation capacity to prevent damage to the duct or piping. Possible configurations and options for allowing displacements while retaining functionality at a
pipe penetration through a floor are shown in Figure 13.4.
Locations where building systems intersect with the building envelope
can require additional considerations. Air intakes can be fit with blast valves
(Figure 13.5) that allow for air flow under normal conditions but close when
excess pressure is detected. These systems can be either passive or active.
Penetration of piping or conduit through the wall can allow blast pressures
(a)
(b)
(c)
Figure 13.4 Examples of Pipe Floor Penetration Supports That Allow Differential
Movement: (a) Pipe riser configurations; (b) Riser guide; (c) Riser isolator (Courtesy
of Kinetics Noise Control, Inc.)
344
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
(a)
(b)
(c)
Figure 13.5 Conceptual Operation of a Blast Valve: (a) Normal flow; (b) Excessive
pressure from outside; (c) Excessive pressure from inside
to enter the building and/or damage the mechanical system. Various available
schemes, including sleeved openings, special accumulators (Figure 13.6) that
allow for pressure buildup and gradual release, and flexible connections can
protect both the piping and the building interior.
Fire protection systems may be considered a subset of piping systems and
designed using the same general principles. However, while the failure of water piping systems is unlikely to jeopardize the building occupants, leaking or
inoperability of fire protection piping (and pipes containing hazardous material)
is potentially life threatening. For this reason it is recommended that additional
caution be exercised when designing these systems.
13.3.9 Electrical Systems
Electric power supply, command, control, and communications systems are
essential to all building systems and a proper threat-assessment matrix evaluation is central to the reliability and survivability of such services. A proper
threat-assessment evaluation consists of identifying critical system components,
dependencies, vulnerabilities, and redundancies, and mapping such items in a
DESIGN CONSIDERATIONS
345
Figure 13.6 Conceptual Design of Accumulator (After U.S. Department of the Army,
1986)
documentable “matrix.” Deliverables resulting from such an effort frequently
result in a Failure Mode and Effects Analysis (FMEA).
For electric systems, certain methods or design techniques are frequently used
in guarding against unconventional threats, including:
r Vertical and horizontal zonal definition of equipment and feeds to provide
redundant paths and feeds
r Self-healing power supply and distribution network designs comprising
multiple, segregated power feeds to minimize human-element intervention
r Command, control, and communication (C3 ) services combining both distributed and centralized control techniques to further minimize the need of
human resources
r Tactical location of critical equipment and personnel
r Co-location of critical equipment in hardened/controlled-access spaces
r Application and integration of automation systems to monitor and orchestrate power, control, and communication systems healing actions
r Appropriate personnel matrix and personnel training to increase the likelihood of survivability/mission continuity, while reducing variable resource
costs
Utility power transformer(s) and switchgear should be located interior to the
building if possible, and in secure spaces. For larger buildings, multiple transformers, separated from each other, enhance reliability should a blast damage
one transformer or its associated equipment. In circumstances where a redundant utility transformer arrangement is not provided, consideration should be
given to locally stocking a replacement transformer and associated equipment
when replacement lead times may be unacceptably long.
346
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
13.3.10 Lighting Systems
Lighting system design for blast protection is similar to electrical system design. The same principles of separation, zoning, emergency power supply, and
so forth, all apply to lighting. One recent development that should be considered
in lighting design is the use of light-emitting diode (LED) technology. More
frequently recognized in the electronic equipment arena, where LEDs are used
for equipment pilot and status lights, LED technology has now evolved to offer
very practical and cost-effective design alternatives in applications where incandescent and fluorescent lamps presently dominate. With respect to susceptibility
to blast damage, LEDs stand alone as a clear leader. In addition, LED lighting reduces the environmental impact of the lighting system and offers reduced
maintenance and increased life.
LED lighting methods do not depend on moving parts or parts that are motionsensitive, as do incandescent and fluorescent lamp technologies (filaments, standoffs, glass, etc); that is, LEDs indeed live up to their designation as “solid-state,”
and their elemental constituents, such as device leads, are solidly embedded into
inert and tough structural polymers. Their substantively smaller sizes and physical profiles minimize concerns related to weight and moment and susceptibility
to blast damage, while also eliminating the need for support devices/equipment
(e.g., ballasts) and the weight/room required for these other support devices.
13.3.11 Other Systems/Considerations
Emergency Power System An emergency generator should be available to provide an alternate source of power if utility power becomes unavailable to support critical life-safety systems such as alarm systems, egress lighting fixtures,
exit signs, emergency communications systems, smoke-control equipment, and
fire pumps. In cases where continuity of critical business missions is necessary, the emergency generator system can also support these missions. Practically, diesel-fueled emergency generator systems are preferred over other conventional fuels (i.e., gasoline, propane, etc.) because of fuel availability, lower
flash point of diesel fuel, and greater power supply capacity from a relatively
small footprint. Economically, maintenance costs are substantially lower for a
diesel-driven emergency generator system when contrasted to other forms of
emergency power, including fuel cells and battery-based systems.
Emergency generators typically require large louvers to allow for ventilation
and cooling of the generator. Care should be taken to locate the generator so that
these louvers are not vulnerable to attack or are fit with blast valves. A remote
radiator system could be used to reduce the louver size and thus reduce air-intake
requirements.
Ideally, emergency power-distribution feeders are contained in hardened enclosures, or encased in concrete, and configured in redundant routing paths to
minimize vulnerability and thus enhance reliability. Emergency distribution panels and automatic transfer switches should be located in hardened spaces isolated
DESIGN CONSIDERATIONS
347
from the normal power system. Emergency lighting fixtures and exit signs along
the egress path could be provided with integral battery packs to provide closetransition services (uninterrupted silhouette illumination) between the time the
utility system fails and the emergency generator system assumes loading.
Emergency Generator System and Fuel Storage A nonexplosive fuel source,
such as diesel fuel, should be considered for emergency generators and dieseldriven fire pumps. The preferred philosophy is to locate all support equipment,
including the diesel fuel day tank, in the same space as the emergency generator,
as a means of minimizing system dependencies. Consideration should be given
to locating the emergency generator day tank in a fire-hardened space adjacent to
the emergency generator space. Day tanks should include sufficient fuel to power
the generator for a minimum of eight hours.
All control equipment, including starting provisions (i.e., air start, battery
start, hydraulic start, etc.), emergency generator switchgear, cooling water tank,
local controls, and so forth, should co-locate with the emergency generator.
Diesel storage tanks that augment the day tanks should be located away from the
building in secure, fire-rated, hardened structures. Fuel piping within the building
should be located in hardened enclosures, and redundant piping systems could
be provided to enhance the reliability of the fuel distribution system.
Fire Control Center A centralized Fire Control Center (FCC) should be
provided to monitor alarms and life-safety components, operate smoke control
systems and fire doors, communicate with occupants, and control the firefighting/evacuation process. Consider providing redundant Fire Control Centers
remotely located from each other, to allow system operation and control from
alternate locations. The FCC should be located near the point of firefighter
access to the building. If the control center is adjacent to the lobby, separate
it from the lobby using a corridor or other substantive buffer area. Provide
hardened construction for the Fire Control Center. Further, the FCC can be
co-located with other critical automated building service functions, such as
with the machinery plant control and monitoring systems. The proliferation of
computer-based monitoring and control systems permits much flexibility in this
regard, particularly for collecting multiple, monitored systems into a single space
through building intranet cabling. If configured, a simple laptop computer can
be used to access and control monitored systems from many possible locations.
Emergency Elevator System Elevators are not used as a means of egress from
a building in the event of a life-safety emergency event, as conventional elevators
are not suitably protected from the penetration of smoke into the elevator shaft.
An unwitting passenger could be endangered if an elevator door opened onto
a smoke-filled lobby. Firefighters may elect to manually use an elevator for
firefighting or rescue operation. A dedicated elevator, within its own hardened,
smoke-proof enclosure, could enhance the firefighting and rescue operation after
a blast/fire event. The dedicated elevator should be supplied from the emergency
348
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
generator, fed by conduit/wire that is protected in hardened enclosures. This
shaft/lobby assembly should be sealed and positively pressurized to prevent the
penetration of smoke into the protected area.
Smoke and Fire Detection and Alarm System A combination of earlywarning smoke detectors, sprinkler-flow switches, manual pull stations, and audible and visual alarms provides quick response and notification of an event.
The activation of any device will automatically start the sequence of operation of smoke control, egress, and communication systems to allow occupants to quickly go to a safe area. System designs should include redundancy
such as looped infrastructure wiring and distributed intelligence such that the
severing of the loop will not disable the system. Install a network of firealarm systems consisting of distributed intelligent fire-alarm panels, such that
each panel can function independently, process alarms and initiate sequences
within its respective zone, while at the same time communicating with other
panels.
Smoke-Control Systems Appropriate smoke-control systems maintain smokefree paths of egress for building occupants through a series of fans, ductwork,
and fire-smoke dampers. Stair pressurization systems maintain a clear path of
egress for occupants to reach safe areas or to evacuate the building. Smokecontrol fans should be located higher in a building, rather than at lower floors,
to limit exposure/access to external vents. Vestibules at stairways with separate
pressurization provide an additional layer of smoke control.
General Announcing Communication Systems A general announcing system
facilitates the orderly control of occupants and evacuation of the danger area
or the entire building. The system is typically zoned by floor, by stairwell, and
by elevator bank for selective communication to building occupants. These systems may be integrated with other building announcing and communication systems. Emergency communication can be enhanced by providing extra emergency
phones separate from the telephone system, connected directly to a constantly supervised central station, and an in-building repeater system for police, fire, and
EMS (emergency medical services) radios.
13.4 LOADING CALCULATION
Nonstructural components can be subjected to several different load types due to
blast. This section details the procedures used to calculate the blast loads, including direct loading, leakage through openings, and propagation through ducts. In
addition, calculation of the building motions for design of systems is addressed,
as is the load due to accidental blastlike loads such as short circuits in electrical
equipment.
LOADING CALCULATION
349
13.4.1 Blast Pressure
Building systems can be subjected to blast pressure from three sources: direct
external blasts, direct internal blasts, and indirect pressures. Equipment located
external to the building shell will be exposed to direct blast pressures. These
pressures are characterized by a short rise time, high peak pressure, and relatively
fast drop of pressure. In contrast, pressures resulting from internal blasts are
confined by the walls and consequently have longer duration. The calculation of
blast pressures due to both external and internal blasts is covered elsewhere in
this handbook (see Chapter 7).
Leakage through Openings Equipment housed in shelters or in rooms with a
direct connection to the building exterior will experience blast pressures from
leakage through openings. These loads may be applied directly to the equipment with connections to the outside, such as air handlers with external intake/exhaust, or indirectly as the pressure fills the room. The calculation of
interior pressure loads due to leakage is complex and usually involves many
assumptions.
One procedure, presented below, incorporates the prominent assumption
that the openings are too small to allow a shock front to develop, and that the
pressure inside the room increases uniformly. The method calculates the average
pressure in the room. The pressure near the opening will be somewhat larger
than the average, but the approximation of uniform pressure within the space is
adequate for design.
The interior pressure loading history is calculated in a step-by-step manner.
The procedure is as follows:
1. Calculate the external pressure-time history.
2. Divide the duration (to ) of the external load into equal intervals of length
δt equal to approximately to /10 or to /20.
3. For each time interval, calculate the change in internal pressure over the
time step, using
δ Pi = C L Ao Vo δt
where CL is the leakage pressure coefficient (Figure 13.7) (Departments of
the Army, the Navy and the Air Force 1990), taken as negative when P is
negative; Ao is the area of the opening; and Vo is the room volume. The
internal pressure at the end of the time interval is then calculated from the
pressure at the beginning of the step as
Pi = P − δ Pi
4. Repeat until the internal pressure drops to the initial ambient pressure.
350
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Figure 13.7 Leakage Pressure Coefficient (After Departments of the Army, the Navy
and the Air Force, 1990)
Example: Calculate the internal pressure-time history for an equipment penthouse with a small opening in one wall.
Given:
Ao = 4 ft2 (2 × 2 opening)
Vo = 1000 ft3 (10 × 10 × 10 penthouse)
Assume peak external pressure at time 0 equal to 3 psi with to = 50 ms.
Solution:
Use δt = to /10 = 50/10 = 5 ms
Perform calculations in a tabular format. Results are shown in Figure 13.8.
LOADING CALCULATION
351
(a)
(b)
Figure 13.8 External and Internal Pressure-Time Histories for Leakage Example: (a)
External pressure-time history; (b) Internal pressure-time history
Propagation of Pressure through Ducts Blast pressure will propagate through
ducts into the interior of a building. The pressure entering the duct, and the
change in pressure at different branching configurations, can be calculated using Figure 13.9. These figures are valid for peak overpressures less than 50
psi, which is appropriate for most civilian structures. In addition, 90◦ bends
in the duct reduce the pressure by approximately 6% each. Thus, after each
352
T (ms)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
...
185
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
P (psi)
Pi (psi)
P-Pi (psi)
CL
δPi
3.00
2.70
2.40
2.10
1.80
1.50
1.20
0.90
0.60
0.30
0
0
0
0
0
0
0.17
0.31
0.43
0.53
0.60
0.65
0.68
0.69
0.69
0.67
0.63
0.59
0.56
0.53
3.00
2.53
2.09
1.67
1.27
0.90
0.55
0.22
−0.09
−0.39
−0.67
−0.63
−0.59
−0.56
−0.53
8.5
7.17
5.91
4.73
3.61
2.55
1.56
0.62
−0.26
−1.10
−1.89
−1.78
1.68
−1.58
−1.49
0.17
0.14
0.12
0.09
0.07
0.05
0.03
0.01
−0.005
−0.02
−0.03
−0.04
−0.03
−0.03
−0.03
0
∼0
—
—
—
bend, the peak pressure, PT , can be calculated from the pressure before the
bend, Po , as
PT = 0.94Po
There is also pressure loss along the length of smooth ducts. However, the
calculation is not suggested for the typical ranges and charge weights assumed
in this chapter, because it is only slightly conservative to assume no loss along
straight portions of ducts. This assumption will be overly conservative for either
long duct runs (over 50 feet) or short ranges. Further details can be found in TM
5-855-1 (U.S. Department of the Army 1986) if more detailed calculations are
desired.
13.4.2 In-Structure Shock
In-structure shock refers to the structural motions that result from a blast. Although equipment located within the structure may not be subjected to direct or
indirect blast pressures, it will experience the resulting structural accelerations.
As a result, building systems not exposed to pressure loads must still be designed
to resist the in-structure shock.
Structural motions due to blast arise from three sources: air-induced ground
shock, direct-induced ground shock, and air shock. Air-induced ground shock
is the result of the pressure wave acting on the ground surface and producing
deflections and hence motion on the ground. Direct-induced ground motions are
due to direct transfer of energy to the ground, typically due to buried or surface detonations. Air shock occurs when the blast pressure impacts the building,
LOADING CALCULATION
(a)
(b)
(c)
(d)
(e)
(f)
353
Figure 13.9 Transmitted Overpressures in Ducts (After U.S. Department of the Army,
1986)
imparting structural motions. The effect of all three sources must be considered
when determining the design motions for building systems.
Air-Induced Ground Shock The peak vertical and horizontal accelerations due
to air-induced ground shock are calculated as
AVai = 100Pso (ρC pg)
AHai = A V tan sin−1 C p (12000U )
where AVai and AHai are the air-induced vertical and horizontal peak ground accelerations (g), Pso (psi) and U (ft/ms) are the peak positive incident pressure and
354
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.2 Typical Values of ρ and Cp (After Departments of the Army, the Navy
and the Air Force, 1990)
Soil Type
ρ (lb-sec2 /in4 )
Soil Type
Cp (in/sec)
Loose, dry sand
1.42E-04
Loose and dry soils
7200–39,600
Loose, saturated sand
1.79E-04
Clay and wet soils
30,000–75,600
Dense, dry sand
1.65E-04
Coarse and compact soils
36,000–102,000
Dense, saturated sand
2.02E-04
Sandstone and cemented soils
36,000–168,000
Dry clay
1.12E-04
Shale and marl
72,000–210,000
Saturated clay
1.65E-04
Limestone-chalk
84,000–252,000
Dry, sandy silt
1.57E-04
Metamorphic rocks
120,00–252,000
Saturated, sandy silt
1.95E-04
Volcanic rocks
120,000–270,000
Basalt
2.56E-04
Sound plutonic rocks
156,000–300,000
Granite
2.47E-04
Jointed granite
9600–180,000
Weathered rock
24,000–120,000
Limestone
2.25E-04
Sandstone
2.10E-04
Shale
2.17E-04
Concrete
2.25E-04
shock front velocity, ρ (lb-sec2 /in4 ) is the soil mass density, Cp (in/sec) is the soil
seismic wave velocity, and g is the acceleration due to gravity (32.2 ft/sec2 ). Typical values of ρ and Cp are given in Table 13.2. If Cp /12000U is greater than 1,
take AHai = AVai .
Direct-Induced Ground Shock Direct-induced ground motions depend upon
the charge weight, distance from the structure, and the soil conditions. The peak
vertical ground acceleration from direct-induced ground shock can be calculated
from
A V di = 10,000 W 1/3 Z G2
where W is the TNT equivalent charge weight (lb), and ZG is the scaled ground
distance (ft). The peak horizontal acceleration depends upon the soil conditions
and is
AHdi =
0.5A V
1.0A V
dry soil
wet soil or rock
The arrival time of the direct-induced ground shock at the structure can be
calculated from
t AG = 12,000RG /C P
where RG is the ground distance from the explosion (ft).
LOADING CALCULATION
355
Air Shock Air-shock motions can be calculated in two ways. A simplified
method is presented in this chapter that assumes a rigid structure and is based
on sliding of the foundation. For this method, only the horizontal motions are
considered, as rocking and vertical motions due to air shock are assumed to
be negligible. Optionally, a more sophisticated analysis can be performed that
accounts for the time history of the load and the flexibility of the structural elements. Although this type of analysis is beyond the scope of this handbook, some
general guidelines are presented. For all but the smallest blast loads, nonlinear
behavior of the structural elements, at least locally near the blast location, is expected and must be included in the analysis model. The load should be based on
the design charge weight and standoff and the time history of pressure applied
to the exposed surfaces of the structure. Care must be taken to ensure that the
stiffness, strength, and mass distribution are modeled as accurately as possible.
The general level of detail used in a dynamic, nonlinear analysis for earthquake
loads is appropriate. One exception to this is that the out-of-plane behavior of
walls is critical in determining the structure response to blast loading, and the
simple linear models used for this behavior mode in earthquake analyses are not
adequate. (See Chapters 6, 7, and 9 for additional information.)
The simplified method starts by assuming that the structure has a single degree
of freedom and that the only resistance to motion is the foundation stiffness. The
resulting equation is
F (t) − FR (t) = M AHas (t)
where F is the applied (blast) load, FR is the resistance of the structure foundation, M is the structure mass, and AHas is the resulting horizontal acceleration.
For a shallow footing, the resistance function can be assumed to depend only
upon friction between the foundation and soil, and is calculated as
FR = µN
where µ is the coefficient of friction (Table 13.3), and N is the normal force
acting on the footings. N is a combination of the structure weight and the vertical
load applied to the roof through the blast pressure. Thus N, and subsequently FR ,
varies with time.
Ideally, the equation of motion would be solved at each time step, and the
maximum acceleration extracted from the results. However, since the load is
only slightly dependent upon the motion, movement in the direction of the load
reduces the effective pressure, and the resistance is assumed to be independent
of motion, a practical alternative is available. The peak pressure and minimum
value of N are assumed, and the resulting acceleration is used as the peak value.
This results in a peak air-shock horizontal structural acceleration of
AHas =
Pso A S − µWs
M
where AS is the exposed area of the structure, and Ws is the structure weight.
356
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.3 Coefficient of Friction between Concrete and Soil (After Departments
of the Army, the Navy and the Air Force, 1990)
Soil Type
µ
Clean sound rock
0.70
Clean gravel, gravel-sand mixture, coarse sand
0.55–0.60
Clean fine to medium or silty medium to coarse sand, silty or clayey gravel
0.45–0.55
Clean fine sand, silty or clayey fine to medium sand
0.35–0.45
Fine sandy silt, nonplastic silt
0.30–0.35
Very stiff and hard residual or preconsolidated clay
0.40–0.50
Medium stiff and stiff clay and silty clay
0.30–0.35
Load Combination The air-induced ground shock and air shock may be assumed to arrive at the structure simultaneously, since both are the result of the
blast pressure. The ground-induced shock wave will arrive at a later time, since
the wave speed through soil is lower than through air. Since only the peak motions are considered, whether or not the motions directly combine depends not
only upon the arrival times, but also upon the actual peak motion times induced
within the structure. The actual time histories of the in-structure motions are
extremely difficult to calculate due to the lack of adequate data on the time
history of the input loads for ground shock and the uncertainty in structural
response.
Whether or not the direct-induced ground shock adds to the air-induced
ground shock and air shock depends upon the arrival time, tAG . If tAG is greater
than the combined arrival time and duration of the blast load (ta + t0 ), then the
air-induced motions can be assumed to dissipate before the direct-induced shock
arrives, and the peak accelerations should not be added. However, if tAG is less
than or equal to ta + t0 , accelerations from all three sources should be combined.
The equations for all three sources of shock were derived assuming a rigid
structure. Based on this assumption, the motion at all levels of the structure will
be identical. However, although structures tend to be very stiff vertically, this
is not true for the horizontal motions. The ground motion will induce horizontal deformations within the structure, resulting in differing levels of acceleration throughout the building height. This variation of in-structure motion with
height has been codified for earthquakes, with accelerations increasing through
the height due to the response of the building. This is not necessarily true for
blast loads, as the short duration does not allow for structural response in the
same manner as earthquake loads. For example, an analysis of a typical threestory steel building subjected to a 10g ground shock load of duration 50 ms, and
accounting for nonlinear structural response, results in absolute accelerations of
10g at the base, 2g at the second story, and 1g at the roof. Relative displacements
between floors are less than 1 inch. Of course, this is only an example, and the
response in other buildings may be vastly different, but it does illustrate the effect
LOADING CALCULATION
357
of the short load duration on building response and the difference with response
to earthquake loads. Thus, great care must be taken when considering the effects
of location within the building, and it is conservative to assume that the ground
accelerations are present at all levels.
The simplified method for calculating air-shock motions does not allow for
differentiation based on location within the structure. Thus, the vertical accelerations in the structure, assuming simultaneous application of the direct- and
air-induced ground shock, are calculated as
A V = AVai + AVdi
and the horizontal ground accelerations are
A H = AHai + AHdi + AHas
Example
Given: Loose, dry sand
ρ = 1.42E − 04
C p = 7200 − 39,600
µ = 0.45
U = 1.4
1000 − lb TNT surface burst @ 100
W = 1000 lb
Pso = 5.5 psi
Z G = 10
Building: 100 wide × 40 tall, assuming typical density gives
Ws = 3,000,000 lb
M = 7764 lb − s2 /in
Solution:
= 5.14g
AVai = 100Pso (ρC pg) = 100(5.5)/[(.000142)(23,400)(32.2)]
AHai = A V tan
sin−1 C p (12000U
=
5.14g
)
AVdi = 10,000 W (1/3) Z 2 = 10,000 (1000)(1/3) (10)2 =10g
G
AHdi = 0.5A V di = 5g
(5.5)(1200)(480) − (0.45)(3,000,000)
Pso A S − µWs
=
= 0.61g
AHas =
M
7764
The total in-structure motions are
A V = AVai + AVdi = 5.14 + 10 = 15.14g
A H = AHai + AHdi + AHas = 5.14 + 5 + 0.61 = 10.75g
358
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Note that these accelerations are much higher than typically seen in seismic events. For earthquakes, the largest accelerations considered in the lower
48 states for nonstructural component design in the 2006 International Building
Code (International Code Council 2006), not including the effect of component
flexibility, are approximately 5g horizontally and 0.6g vertically.
Anchorage Force Calculation The anchorage forces are calculated by applying
the shock load at the equipment center of mass. The applied force is equal to
the shock acceleration multiplied by the equipment mass. The force should be
applied in all directions, unless the possible blast locations limit the shock direction. Usual practice is to also consider eccentricity of the center of mass, as well
as the actual attachment point locations.
Seismic design principles require that the applied load be further modified
based on the flexibility of the equipment and the amount of nonlinear behavior
that can be tolerated before damage. The flexibility component is intended to
magnify the forces if the equipment will move significantly relative to the floor
(increasing the effective force). For two pieces of equipment of identical mass,
the more flexible equipment will undergo larger deformations and experience
larger accelerations at its center of mass. Another factor that can modify calculated forces recognizes that most equipment can undergo some nonlinear behavior without sustaining significant damage. Since the analysis used to evaluate the
equipment and anchorage assumes linear behavior, a lower effective force will
be transmitted during the actual event. The amount of the reduction that is used
depends on the ability of the equipment to accept nonlinear deformations without damage. However, no formal guidance exists regarding the applicability of
these factors for blast loading, and great care should be taken if they are to be
applied for blast analysis.
Example: Motor Control Center
Equipment data (Figure 13.10):
W
A
B
H
= 2800 lb
= 80 in
= 20 in
= 92 in
Load Data: Use values from previous examples.
A V = 15.14g
A H = 10.75g
Pso = 5.5 psi
Shock calculation:
r Apply the loads AV and AH at the center of mass (assume at center in plan
and midheight). Account for 5% eccentricity in any direction.
(a)
(b)
Figure 13.10 Example Calculation Equipment Layout: (a) Motor control center; (b)
Anchorage layout
360
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
r Calculate the anchorage forces assuming that static loads equal to AV W and
AH W are applied simultaneously. The horizontal load can be applied in any
direction with the maximum anchorage forces governing.
The applied loads are:
Vertical = 15.14g ∗ 2800 lb = 42,392 lb (uplift or compression)
Horizontal = 10.75g ∗ 2800 lb = 30,100 lb
The anchorage forces with the horizontal load applied in the short direction with
zero eccentricity are:
Uplift = (42,392 − 2800)/4 + (30,100)(46 in)/[2(20)] = 44, 513lb
Compression = (42,392 + 2800)/4 + (30,100)(46 in)/[2(20)] = 45, 913lb
Horizontal = 30,100/4 = 7525lb
and the resulting maximum anchorage forces considering eccentricity are:
Vertical = 47,632 lb uplift
Horizontal = 8455 lb
Note that these are considerably higher than 1/4 of the applied loads due to
the location of the load application and the assumed eccentricities. In particular,
the vertical forces are greatly increased due to the overturning moment imposed
by the horizontal force.
These anchorage load values exceed the capacity of 11/4 anchors by approximately 25%. Adequate anchorage would require multiple anchors at each support
location. This compares to one 1/2 anchor required at each support location in
an area of moderate seismic activity (UBC, Zone 3) (International Code Council
1997).
Blast Calculation: Three basic approaches can be used to apply blast loads to
equipment. First, the peak (or average) blast pressure can be applied as a static
force. This is an extremely conservative approach for most equipment, as it ignores the short duration of the applied load. The second approach is to consider
the blast load to be impulsive in nature and to use a simplified method to determine the maximum response. A drawback to this approach is that the dynamic
characteristics, mass and stiffness, of the equipment must be known, which is
usually not the case. A third alternative is to explicitly calculate the time history
response of the equipment to the blast load, but this approach is rarely justified
due to the—at best—approximate nature of the input parameters and the requirement to know the equipment mass and stiffness distribution.
The first two approaches will be considered in this example.
Method 1 – Static Load Application
LOADING CALCULATION
361
r Apply the pressure as a static load distributed evenly across the exposed
face.
r Calculate the anchorage forces for the given loads. The horizontal load can
be applied in any direction, but for symmetric equipment, applying the load
along each principal axis is usually sufficient.
The equivalent applied forces are:
Strong axis = 92 in ∗ 20 in ∗ 5.5 psi = 10,120 lb
Weak axis = 92 in ∗ 80 in ∗ 5.5 psi = 40,480 lb
The resulting maximum anchorage forces are:
Strong axis loading vertical = 5119 lb (uplift)
Strong axis loading horizontal = 2530 lb
Weak axis loading vertical = 185,508 lb (uplift)
Weak axis loading horizontal = 10,120 lb
The anchorage forces generated by these loads are clearly beyond the capacity
of a cost-efficient anchorage system. Several simplifying assumptions were made
in this calculation. In addition to the static application of the load, no load was
applied to the top, side, or rear surfaces. Loads applied on the top and rear face
will reduce the anchorage forces.
The exposed front face of the motor control center must be able to resist the
5.5 psi load. The ability of most sheet metal to sustain such a load is doubtful
at best, with maximum duct pressures approaching 1 psi. Thus, although this
example demonstrates the procedure for calculating the anchorage forces to a
load of 5.5 psi, the equipment would have to be protected through barriers or
relocation to reduce the load to a level that could be resisted.
Method 2 – Impulsive Load
A typical natural frequency for lateral motion of a motor control center might
be 0.5 sec. For this example, that is a lateral stiffness of roughly 1000 lb/in. From
Figure 13.11 the dynamic amplification factor for a ratio of
0.050
t0
=
= 0.1
T
0.50
is D = 0.35. This leads to a maximum total horizontal reaction force of
Htotal =
3542 lbs
0.35(5.5)(92)(20)
=
0.35(5.5)(92)(80)
14,168 lbs
for the strong and weak directions, respectively. This calculation yields horizontal support loads of 885 and 3542 lbs, respectively. Similarly, the vertical forces
362
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Figure 13.11 Dynamic Amplification Factor for Triangular Impulse Loads
due to blast are D times the static values. This results in
V =
2246 lbs
65,382 lbs
The vertical loads (uplift) are still very large, but are likely able to be resisted
with proper anchor design, certainly if direct attachment to steel is possible.
Stiffer equipment (smaller T) will have larger dynamic amplification factors,
while more flexible equipment will see greater reductions in force.
13.5 SUMMARY
Presently, there is little guidance and few standards for blast-resistant design of
building systems for civilian structures. Although there are several efforts currently under way, familiarity and selective adoption of methods used in other
arenas, such as used by the DoD in hardening combatant-type systems, can be
economically and practically exported to the civilian building theater. Likewise,
adoption of new and available technologies, such as LED lighting, can provide
reasonable stopgap measures.
This chapter has attempted to highlight the main areas of concern, provide
basic guidance as to the issues to be addressed, and demonstrate computational
techniques where appropriate. However, given the lack of recognized, standardized design procedure, it is imperative to recognize that, as in all engineering,
there is no substitute for common sense and engineering judgment. The
REFERENCES
363
principles of design remain unchanged, despite the lack of exact knowledge
regarding the loading and capacity for resistance.
REFERENCES
American Society of Heating, Refrigerating and Air-Conditioning Engineers. 2007.
ASHRAE Handbook: Applications. Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers.
Departments of the Army, the Navy and the Air Force. 1990. Structures to Resist the
Effects of Accidental Explosions, rev. 1 (Department of the Army TM 5-1300, Department of the Navy NAVFAC P-397, Department of the Air Force AFM 88-22).
Washington, DC: Departments of the Army, the Navy and the Air Force.
International Code Council. 1997. Uniform Building Code. Washington, DC: International Code Council.
. 2006. International Building Code. Washington, DC: International Code
Council.
National Fire Protection Association. 2005. National Electrical Code Requirements for
One- and Two-Family Dwellings (NFPA 70A). Quincy, MA: National Fire Protection
Association.
U.S. Department of the Army. 1986. Fundamentals of Protective Design for Conventional Weapons (TM 5-855-1). Washington, DC: U.S. Department of the Army.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
IV
Blast-Resistant Detailing
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
14
Blast-Resistant Design
Concepts and Member
Detailing: Concrete
Steven Smith and W. Gene Corley
The next four chapters provide design and detailing guidelines for members
constructed from reinforced concrete, steel, and masonry, as well as members
strengthened using fiber-reinforced polymers (FRP).
The general approach for all of these materials systems is to use two independent approaches to providing the toughness that is critical to blast resistance. The
approaches are informally associated with strength and ductility (which, taken
together, provide toughness). Strength is scaled through the response limits for
each level of protection (LOP). For example, the response limit targets for LOP
IV are elastic response. Thus, members designed to this LOP will be larger and
stronger than those designed to lower LOP (which allow greater displacement).
However, recognizing the universal need for ductility, the detailing recommendations are often constant for all LOP. For the lower LOP, this provides the ductility
necessary to achieve the expected response limits. For higher LOP, this detailing provides a higher margin against the uncertainties inherent to blast-resistant
design.
14.1 GENERAL
14.1.1 Scope
Detailing recommendations for reinforced concrete structures are provided in
this chapter. The nature of high-rise structures in a dense urban environment
introduces significant complexity, including near-field blast loads and intricate
structural members such as transfer girders. For this reason, the methods presented in this section, while generic in nature, are best applied to low-rise structures with moderate standoff distance (say, Z > 3.0).
The seismic provisions in ACI 318-08, Building Code Requirements for Structural Concrete (American Concrete Institute 2008), provide the basis for the detailing, although many references are also made to UFC 3-340-02, Structures
367
368
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
to Resist the Effects of Accidental Explosions (Unified Facilities Criteria 2008).
Chapter 21 of ACI 318-08 contains requirements for design and construction of
structures subjected to earthquake motions. The requirements are scaled to the
severity of earthquake motions and are identified as ordinary, intermediate, and
special moment frames. Special moment frame (SMF) requirements provide the
highest level of performance and are the primary resource for the ACI provisions
referenced in this chapter.
The general intent of seismic detailing is to provide greater ductility and
post-yield strength than conventional detailing. This is consistent with the blastresistant design intent of improving toughness. The Portland Cement Association’s Blast Resistant Design Guide for Reinforced Concrete Structures (Portland
Cement Association 2009) contains additional discussion of structural aspects of
seismic and blast phenomena, as do Hayes (Hayes et al. 2005) and Bilow (Bilow
and Kamara 2007).
14.2 FAILURE MODES
14.2.1 Flexural
Flexure is in many instances intended to be the primary structural mode for members resisting blast loads. This is due to two related concepts: the relatively predictable nature of, and the ductility associated with, flexural response.
Columns subjected directly to lateral air-blast pressures are often the single
most critical design consideration for blast-resistant design. Columns directly
exposed to blast pressures will respond primarily in flexure. In fact, UFC
3-340-02 recommends disregarding the column axial load and analyzing it
solely as a beam. Neglecting the compressive preload is generally a conservative
approach to sizing columns, as long as a separate stability check is performed
that includes the axial load.
Columns may develop a three-hinged mechanism with plastic hinge zones
located at end supports and at an intermediate point along the length of the column. The location of the intermediate plastic hinge can be determined by analysis and depends on the location of the blast source as well as the height and
properties of the column. However, this analysis can be complex and may require assumptions that render results of little use. For most situations, the best
approach is either to conservatively design the column for the potential of a hinge
anywhere along the entire length, or to make general assumptions of intermediate hinge location based on the anticipated blast loading. If the column is in
the far field, the blast pressure will be essentially uniform, and the intermediate
hinge will form near the center of the column (assuming no other structural details that would alter the symmetry of the column response). If the column is in
the near field, and subjected to blast from a package bomb, the blast pressures
will be concentrated at the base of the column, and move the intermediate hinge
downward.
FAILURE MODES
369
Seismic design philosophy as applied in FEMA 350, Recommended Seismic
Design Criteria for New Steel Moment-Frame Buildings (Federal Emergency
Management Agency 2000) employs a target hinge rotation of 0.03 rad (approximately 2◦ ). This hinge rotation can also be associated with reinforced concrete
design (Bilow and Kamara 2007). Potential hinge rotations under blast loading,
however, span a wide range. as described elsewhere in this handbook. The
target seismic deformation magnitude, and overall response intent, would be
best associated LOP II–III. The extreme deformations, structural damage, and
allowance of tension membrane associated with LOP I are beyond the intent of
seismic design.
14.2.2 Diagonal Tension
Diagonal tension shear is a brittle failure mode and must be avoided. Care
must be taken to provide sufficient shear capacity to allow for the full flexural development of the section. The phi-factors provided in ACI 318-08 are
intended to promote preferential failure modes; a greater strength reduction is
applied to shear than flexure. However, the structural analyst has the task of
obtaining the correct member reactions, particularly when conducting singledegree-of-freedom (SDOF) analysis and mapping the results back to the physical
system.
14.2.3 Direct Shear
The primary focus of blast analysis is typically the dynamic flexural response of
members. However, earlier during the blast event, direct shear forces are transmitted to the element supports. The magnitude of the direct shear force may be
significantly greater than the reaction due to dynamic flexural response, though
the resistance to direct shear is also typically greater that diagonal shear resistance.
Direct shear can be an even more brittle mode of response than diagonal tension. Direct shear is the phenomenon of failure at a shear plane, typically at the
face of a support. Direct shear is resisted by shear friction and inclined reinforcement. Specific detailing guidelines for beams and columns are provided later in
this chapter.
14.2.4 Membrane
Slabs and beams that develop extreme deformations from large, directly applied
blast loading may be designed to transition from a flexural response to a tension
membrane response. In addition to careful detailing of the slab or beam, this
mode of response requires careful consideration of the large in-plane forces that
will be transmitted to the surrounding structural elements.
370
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
14.2.5 Stability
Stability is a critical design factor for columns. Designing the column to support
some or even the entire design load after loss of intermediate lateral support is
an indirect means of providing substantial buckling resistance. This equates to
a scenario of the loss of bracing from floor framing, and essentially requires
designing the column for an unbraced length equal to two story heights.
This is not a blast-resistant design method per se, although it clearly relates
to the way the building will resist progressive collapse. However, this is such
an important design consideration that it should be carefully considered, even if
formal progressive collapse design is not utilized. And, this is an essential design
concept whenever a strong-column, weak-beam approach is used.
14.3 DETAILING
14.3.1 General
General procedures that will promote toughness include:
1. Use of concrete with a minimum strength greater than 3,000 psi. This will
help to ensure against brittle shear failure.
2. Concrete with strength greater than 10,000 psi should be used with caution
due to its potentially brittle failure mode in flexure.
3. All concrete should be normal weight. Existing test data and analysis of the
response of reinforced concrete to blast and seismic loading are primarily
based on normal-weight concrete. Thus, for normal-weight concrete, concepts such as strain-rate effects, fragmentation, and pressure-impulse (P-I)
calibration for deformation limits are best understood.
4. The ACI 318-08 design compressive strain in concrete of 0.003 should also
be applied to blast-resistant design, to avoid crushing of concrete and an
associated brittle failure in flexure.
5. Reinforcing bars are recommended to be ASTM A706 (ASTM 2006).
Where ASTM A615 (ASTM 2008) bars are used, limitations as specified
in ACI 318-08, Section 21.2.5, should be applied. The primary concerns
for reinforcing steel are the ratio of yield to ultimate strength and the actual yield versus design yield strength. A large ratio of yield to ultimate
strength is necessary to provide sufficient ductility and energy absorption.
Tests indicate that reinforcing that has actual tensile strength exceeding
the actual yield strength by at least 25% provides substantial yield region.
Additionally, the actual yield strength should not exceed the design yield
strength by more than 18,000 psi. The intent of this limitation is to prevent
the development of undesirable failure modes (such as shear) prior to the
formation of plastic hinges.
6. Reinforcing bar sizes should be smaller than No. 11.
DETAILING
371
7. Straight reinforcing bars should be used to avoid reduced ductility at bends.
8. Continuity with In-plane Members: Every opportunity should be taken to
provide a load path from framing members to plane elements. For example,
beams should be tied to floor slabs, and columns to bearing walls. Particularly in situations when blast forces would be directed against the weak
axis of the member (often the case for beams), this approach can provide
the required blast resistance without having to substantially increase the
dimensions of the member.
9. Balanced Design: Where appropriate, members should be designed to resist the full capacity of supported elements (as defined by a governing
failure mechanism), rather than the design loads transmitted by those elements. For example, it may be desirable for a girder supporting a floor
slab to be designed to resist in-plane forces developed by catenary action
in the slab, even though these forces might be significantly greater than
forces developed from the design loads alone with the slab responding in a
flexural mode.
10. Weak-Beam, Strong-Column: Beam failure is generally preferable to column failure, since the beam failure typically only affects the structure in
close proximity, whereas the column failure could lead to progressive collapse.
11. ACI 318-08 includes provisions for spiral reinforcement (referred to as
“laced” in UFC 3-340-02). Spiral reinforcement provides confinement and
torsion and shear capacity that are superior to stirrups and rectangular hoop
reinforcement. References to selected ACI 318-08 guidelines for spiral reinforcement are included below. UFC 3-340-02 includes more detailed
information on detailing members with laced reinforcement. In general,
the higher performance of spiral reinforcement is most necessary for very
close-in blasts (Z < 1.0), which are beyond the scope of this guide. But
any situation requiring a high level of structural performance merits consideration of spiral reinforcement.
14.3.2 Splices
Splicing can be achieved by lapping bars (No. 11 bars and smaller), mechanical
connectors, or welding. It is good practice to locate splices in noncritical regions,
where low stress states are expected under blast conditions. In particular, splices
should be avoided in locations where plastic hinges are likely to form and at
member intersections. The percentage of bars spliced in a single location should
be limited to 25 percent if possible. Splices should also be separated along the
axial length of a member by at least twice the member depth (Figure 14.1).
Lap splices should satisfy ACI 318-08 Class B requirements, which provide
for a 30% increase in development length over Class A lap splices. Lap splices
should be enclosed by lateral reinforcement conforming to Sections 21.6.4.2,
21.6.4.3, and 21.6.4.5 of ACI 318-08, to provide adequate confinement of the
372
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
Abfy
Abfy
1.3 ld
1.3 ld
1.3 ld
(min)
1.3 ld
2 d (min)
1.3 ld
Figure 14.1 Schematic of ACI 318-08 Lap Splice Spacing and Grouping Recommendations.
concrete surrounding the splices, and to protect the splices from the effects of
spalling. Noncontact lap splices are not recommended. According to UFC 3340-02, Section 4-21, splices of adjacent bars should be staggered by at least the
required lap length, since overlap of splices of adjacent bars is undesirable. Section 21.5.2.3 of ACI 318-08 advises that lap splices in flexural members should
not be used within joints (a) within a distance of twice the member depth from
the face of a joint, or (b) at regions of flexural yielding. Since hinge locations
can be difficult to anticipate for blast loading, welded and mechanical joints are
preferred for blast-resistant design.
If mechanical splices are used, they should satisfy ACI 318-08 Type 2 requirements, which call for at least 125% of the specified yield strength of the bar. UFC
3-340-02 necessitates testing to exhibit the adequacy of mechanical splices under dynamic conditions prior to acceptance for use in hardened structures. UFC
3-340-02 requires that mechanical splices develop both equivalent tensile capacity and ductility of a continuous bar; however, no mechanical splices have been
shown to satisfy this ductility requirement. Thus, where ductility is of paramount
importance, mechanical splices should be avoided.
UFC 3-340-02 suggests that welded splices should be avoided, since welding
may result in a reduction of the ultimate strength and ductility of the reinforcing
steel. However, if welding is necessary, strict adherence to the requirements
of ANSI/AWS D1.4, Structural Welding Code—Reinforcing Steel (American
Welding Society 2005), should permit proper development of the reinforcing
steel strength. Use of testing to demonstrate the performance of welded
reinforcing steel is also good practice.
14.3.3 Columns
Columns subjected directly to lateral air-blast pressures are often the most critical single design consideration. These columns experience large shear forces,
DETAILING
373
out-of-plane deformations that can lead to instability, and reflected pressures that
may have the potential to shatter concrete.
The flexural response of the columns may develop a three-hinged mechanism
with plastic hinge zones located at end supports and at an intermediate point
along the length of the column. The location of the intermediate plastic hinge
can be determined by analysis and depends on the location of the blast source as
well as the height and properties of the column. This analysis can be challenging,
and may require assumptions that render results of little use. For most situations,
the best approach is either to conservatively design the column for the potential
of a hinge anywhere along the entire length, or to make general assumptions of
intermediate hinge location based on the anticipated blast loading. If the column
is in the far field, the blast pressure will be essentially uniform, and the intermediate hinge will form near the center of the column (assuming no other structural
details that would alter the symmetry of the column response). If the column is
in the near field, and subjected to blast from a package bomb, the blast pressures
will be concentrated at the base of the column, and move the intermediate hinge
downward.
The P-Delta moment generated by the inelastic response must be addressed.
Of critical concern is whether the column can develop inelastic response and
maintain sufficient moment capacity to resist the associated P-Delta effect.
FEMA-350, Recommended Seismic Design Criteria for New Steel MomentFrame Buildings (Federal Emergency Management Agency 2000), addresses
this concern for structural steel seismic special moment frames by requiring
the moment capacity at 0.03 radians of inelastic response to be greater than
the nominal plastic moment, Mp . A similar rationale can be adopted for reinforced concrete members, and a synthesis study of available reinforced concrete SMF data indicates that this detailing provides inelastic response of 0.03
radians at the connections. This level of response would equate to the Moderate to Heavy criteria which are typically associated with LOP III and II,
respectively.
Detailing of beam-column joints is discussed in Section 14.3.5 of this chapter,
and per the above discussion it is an essential component of achieving P-Delta
resistance by providing lateral restraint at frame points.
P-Delta resistance within the clear height of the column is achieved with
proper reinforcement detailing so that inelastic response in this region does not
unacceptably reduce the flexural resistance of the column. Detailing for columns
should be based on SMF provisions in ACI 318-08, Section 21.6. The following
subsections present the primary issues to address.
Longitudinal Reinforcement Section 21.6.3.1 of ACI 318-08 limits the area of
longitudinal reinforcement to between 1% and 6% of the gross area. The lower
limit provides that the column will achieve the yield moment before the cracking
moment. The upper limit addresses congestion and avoiding development of high
shear stresses. To minimize the number of iterations, it is typically best to begin
the design process with reinforcing steel at the 6% limit.
374
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
Transverse Reinforcement (Diagonal Shear and Confinement) Transverse reinforcement serves the equally important purposes of resisting diagonal tension
and providing confinement. Confinement from closely spaced hoops protects the
core concrete from being shattered by high, directly applied blast pressures, or
crushed by large deflections and stresses at locations of large inelastic flexural
response. Maintaining the integrity of the core concrete is critical for lateral restraint of the longitudinal reinforcement and allows the concrete to contribute
to the shear capacity of the section. Hoops and ties for confinement are particularly critical in regions of inelastic response. Per the previous discussion of
intermediate hinge location, unless that position can be reliably predicted, it is
recommended that transverse reinforcement be provided per Section 21.6.4 of
ACI 318-08 over the clear height of the column, rather than only over a distance
l0 , as defined in Section 21.6.4.1. Where steel jacketing is provided, transverse
reinforcement can be reduced, since the steel jacket will protect against spalling
and provide the necessary confinement.
Transverse reinforcement spaced to satisfy confinement requirements (Section 21.6.4) will likely satisfy shear requirements for diagonal tension; however,
the requirements of Section 11.4 should always be checked, particularly when
the confinement requirements of Section 21.6.4 are not provided over the clear
height of the column.
Direct Shear Columns must be designed to resist the effects of direct shear as
well as flexure and diagonal tension, although in most cases the direct shear will
not control.
Resistance to direct shear from blast loading can be treated in a manner similar
to that of Section 11.6 of ACI 318-08, which provides minimum requirements for
shear friction reinforcement. The shear friction concept presumes a crack at the
critical shear plane, such that sufficient reinforcing steel is required to maintain
interlock of the rough, mating surfaces. Since the function of the shear friction
reinforcement is to resist the tendency of the shear plane to open, it is the tensile
capacity (not shear) that is of interest; as long as there is sufficient steel, and
it does not yield, the mating surfaces will remain engaged and resist the shear.
Presuming sufficient shear friction reinforcement (as defined ACI 319, Section
11.6), the direct shear capacity of the concrete and any inclined bars that cross
the shear plane can be calculated as
Vn = 0.18 f cd bd + As f yd cos α
Where, f cd
is the product of the concrete compressive strength, DIF and SIF; As
is the area of inclined steel bars; f yd is the product of the steel yield stress, DIF
and SIF; and α is the angle formed by the plane of the inclined reinforcement and
the surface of the shear plane. Note also that since the direct shear phenomenon
occurs very early during the blast application, and before the development of
flexural response, Type I cross section is used. Type I cross section is defined as
the uncracked cross section; that is, the complete concrete cross section can be
counted for resisting the direct shear.
DETAILING
375
7
6
5
KEY
3
4
3
1. Steel jacket – continuous below grade
2. No splice within 2d of steel jacket
3. Splice zone – expected linear behavior
(Type B lap or type 2 mechanical)
4. Hinge zone – no splice
5. No splice within 2d joint
6. Transverse reinforcement continued through joint
7. Transverse reinforcement spaced per ACI 318-08
21.6.4 over entire length (or use round, spiral
reinforced column)
2
1
Figure 14.2 Schematic of Column Detailing Recommendations
Splice Locations Even though hinge location can be difficult to estimate, as
discussed previously, the attempt should be made so that splices can be avoided
at those locations. Splices should never be located in a region that does not have
transverse reinforcement placed per Section 21.6.4 of ACI 318-08.
Additionally, splices should not be located within a joint or within a distance
of twice the member depth from the face of a joint.
Unbraced Length Designing the column to support some or all of the design
load after loss of intermediate lateral support is an indirect means of providing
substantial buckling resistance. This equates to a scenario of the loss of bracing from floor framing, and essentially requires designing the column for an
unbraced length equal to two story heights.
This is not a blast-resistant design method per se, although it clearly relates
to the way the building will resist progressive collapse. However, this is such
an important design consideration that it should be carefully considered, even if
formal progressive collapse design is not utilized. And, this is an essential design
concept whenever a strong-column, weak-beam approach is used.
14.3.4 Beams
This section addresses beams with clear spans greater than four times the beam
effective depth. Beams shorter than this should be treated as deep beams with
shear capacity determined by methods in other sources. Use of shear beams for
376
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
blast-resistant design should be carefully considered due to the potentially brittle
failure modes associated with them.
Exterior beams subjected directly to lateral air-blast pressures experience
large shear forces, out-of-plane deformations that can lead to the development
of plastic hinges, and reflected pressures that may have the potential to shatter
concrete. Interior beams may be subjected to either direct blast pressures in the
form of uplift (for situations in which the interior of the building becomes pressurized) and forces transmitted from other structural members.
Plastic hinge development for exterior beams is similar to that for columns,
as previously discussed. Plastic hinge zones can develop at fixed supports and at
intermediate points along the length of a beam. The exact location of an intermediate plastic hinge can be determined by analysis and depends on the location
of the blast source as well as the span and properties of the column. In lieu of
such complex analysis, two general cases can be considered to bound the problem. Beams subjected to far-field blast load will most likely develop hinges at
supports and at midspan. This assumes that the load applied to the beam from
a far-field blast will be nearly uniform over the column length. Sources of exceptions to this assumption could include very long span beams and beams with
intermediate framing. Beams subjected to near-field blast load will most likely
develop the middle plastic hinge at a location biased to the position of the blast
source.
Simply supported beams can be used in blast-resistant structures if they can be
shown to satisfy the required level of performance. However, the designer must
be aware that the simple supports limit the available potential energy dissipation
and additional margin against unanticipated blast effects.
Detailing for beams should be based on SMF provisions in ACI 318-08, Section 21.5. The following subsections present primary issues to address during the
design process.
Longitudinal Reinforcement Longitudinal reinforcement should comply with
ACI 318-08, Sections 21.5.2.1 and 21.5.2.2. Beams subjected to uplift should
have reinforcement designed to resist both the uplift and rebound.
Transverse Reinforcement (Diagonal Shear and Confinement) Transverse reinforcement serves the equally important purposes of resisting diagonal tension
and providing confinement. Confinement from closely spaced hoops protects the
core concrete from being shattered by high, directly applied blast pressures, or
crushed by large deflections and stresses at locations of large inelastic flexural
response. Maintaining the integrity of the core concrete is critical for lateral restraint of the longitudinal reinforcement, and allows the concrete to contribute
to the shear capacity of the section. Hoops and ties for confinement are particularly critical in regions of inelastic response. Transverse reinforcement should be
spaced to satisfy 21.5.3.2 and provided over the clear span of the beam, unless
specific analysis of yield zone location supports alternate placement. Transverse
reinforcement spaced to satisfy 21.5.3.2 will likely satisfy shear requirements
DETAILING
377
for diagonal tension; however, the requirements of Section 11.4 should always
be checked, particularly when the confinement requirements of Section 21.5.3.2
are not provided over the clear span of the column.
Single-piece hoops as shown in Detail A of Figure R21.5.3 of ACI 318-08 are
preferred to hoops built up from multiple members.
Direct Shear The discussion and recommendations for direct shear in columns
can be directly applied to beams.
Splice Location Even though hinge location can be difficult to estimate, as
discussed previously, the attempt should be made so that splices can be avoided
at those locations. Splices should never be located in a region that does not have
transverse reinforcement placed per Section 21.5.4 of ACI 318-08.
14.3.5 Beam-Column Joints
Joint design is critical to providing an engineered response to blast loading. The
development of inelastic response, which typically accounts for the majority of
energy dissipation for LOP I–III, is directly related to connection performance.
The performance intent of LOP IV is essentially elastic response to the design
blast load; achieving this intent, and also providing margins against unanticipated
load magnitudes, requires equally careful joint detailing.
The performance expectations of the various levels of protection can potentially be achieved with a simple joint between beams and columns, but the tradeoff of this design approach is potentially massive beams.
The monolithic nature of reinforced concrete can be used to great advantage in
the development of robust beam-column joints. Reinforced concrete joints inherently provide substantial deformation continuity between beams and columns.
To provide the level of performance demanded for blast resistance, measures
must be taken to ensure ductility and to protect the core concrete. These goals
are primarily achieved through detailing considerations for the beam longitudinal reinforcement (ductility) and column transverse, confinement reinforcement
(protection of core concrete).
Longitudinal Reinforcement Wherever possible, beam longitudinal reinforcement should be continuous through joints. Splices should not occur within joints,
or within a distance of twice the beam depth from the face of the joint.
Reinforcement that terminates at a joint should comply with Section 21.7.2.2,
which requires the reinforcement to be anchored at the far face of the confined
column core.
Transverse Reinforcement Column transverse reinforcement should be continued through the joint, as described in Section 21.7.3. This section provides
consideration of the similar function that is provided by continuous beam longitudinal reinforcement for joints that have beams framed to all four faces. The
378
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
6
5
4
1
2
3
2
1
KEY
1. No splice w/in 2d from column face
2. Splice zone – expected linear behavior
(Type B lap or type 2 mechanical)
3. Hinge zone – no splice
4. Standard hook for bottom and top reinforcement
(to resist uplift)
5. Longitudinal reinforcement continuous through joint
6. Transverse reinforcement spaced per ACI 318-08
21.5.3 over entire span
Figure 14.3 Schematic of Beam Detailing Recommendations
important consideration is to have confining reinforcement around the column
longitudinal reinforcement within the joint.
Shear Strength Shear strength should be computed per Section 21.7.4. Also of
importance in this section is the definition that a beam is considered to provide
confinement to the joint if a least three-quarters of the face of the joint is covered by the beam. This definition applies to the calculation of shear resistance.
However, relative to the consideration of transverse reinforcement discussed previously, the most important consideration is enclosure of the column longitudinal
reinforcement that continues through the joint. Thus, a beam that satisfies 21.7.4
for shear capacity of the joint should not necessarily negate the use of transverse
reinforcement, spaced as previously discussed, to provide additional protection
of the joint core.
14.3.6 Slabs
The designer must first determine the structural behavior anticipated for the slab.
Slabs can serve numerous roles, including:
Flexural element: This is a conventional role for a building slab. With respect
to blast resistance, a slab will respond as a flexural element to directly applied
blast forces. This can occur with explosions initiated within the building, or
when the interior becomes pressurized by failure of exterior elements. Blast
pressures acting on the soffit of the slab will generate uplift and stresses opposed
to gravity loads.
DETAILING
379
Diaphragm: Slabs that are intended to act as diaphragms for lateral response
to conventional loads can also be used to efficiently distribute blast forces
through the structure, and thus dissipate the blast effects by engaging additional
structural elements and building mass.
Local Restraint: Slabs that are not explicitly intended to perform as a diaphragm can still be used to provide local lateral restraint to beams that are
directly exposed to transverse blast forces. This can be a very effective way to
support beams against the weak-axis bending.
Membrane: Slabs that develop extreme deformations from large, directly applied blast loading may be designed to transition from a flexural response to
a tension membrane response. In addition to careful detailing of the slab, this
mode of response requires careful consideration of the large in-plane forces the
slab will transmit to the surrounding structural elements.
The following subsections are concerned with slab-detailing considerations.
Flexural Reinforcement The flexural reinforcement ratio should not be less
than 0.0018. This is the limit for temperature and shrinkage, when using 60 ksi
reinforcement, given in Section 7.12. In practice, the reinforcement ratio will
almost always exceed this.
Flexural reinforcement should be provided at top and bottom throughout the
slab. For blast scenarios in which uplift is expected, the top reinforcement should
be explicitly designed to resist the uplift forces, and the bottom reinforcement to
resist rebound. In the event that design blast loads will not subject the slab to
uplift, but the slab is part of the structural system that will resist blast effects,
it is good practice to provide approximately equal amounts of top and bottom
reinforcement. This approach will provide reserve capacity for events such as an
unanticipated reversal of curvature or loss of a support.
Bottom reinforcement should be continuous across interior supports. This is
particularly important in the column strip of two-way flat-slab floor systems.
At exterior supports, the reinforcement should be fully developed. This bottom
reinforcement detailing is a precaution against the loss of a support, providing
the structural potential for the slab to bridge across the lost support.
Connected Elements Structural elements connected to a slab should generally
be designed to establish the full capacity of the slab (i.e., balanced design). This
must be carefully considered where the slab is intended to develop tension membrane behavior, because the in-plane forces that will need to be equilibrated by
the surrounding structure will be many times larger than the flexural forces transferred from the slab.
Slabs that serve as lateral restraint for beams directly subjected to blast pressures must have sufficient connectivity for the beam to engage the mass and
in-plane stiffness of the slab.
Fully distributed force transfer between the slab and connected elements is
the preferred mechanism for all situations. For structures with concrete framing
members, this is readily achieved. With steel-framed structures, the best method
380
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
is to cast the member into the concrete slab. For example, for a steel girder supporting a concrete deck, the top flange of the girder should be cast into the deck,
if possible, rather than using shear studs.
14.3.7 Walls
As with slabs, the designer must first determine the structural behavior anticipated for the wall. Walls can serve numerous roles, including:
Gravity System: As part of the gravity load system, walls may experience
uplift from internal pressurization, or, in rare instances, increased compressive
loads from internal overpressures at higher floors.
Flexural Element: A wall will have to be able to respond in flexure in the event
of blast forces applied directly to one face, or unbalanced blast forces applied to
both faces. This can occur with explosions that are set off within a building, or if
the interior becomes pressurized from failure of exterior elements.
Diaphragm: Walls that are intended to act as diaphragms for lateral response
to conventional loads can also be used to efficiently distribute blast forces
through the structure, and thus dissipate the blast effects by engaging additional
structural elements and building mass.
Local Restraint: Walls that are not explicitly intended to perform as a diaphragm can still be used to provide local lateral restraint to columns that are
directly exposed to transverse blast forces.
The following subsections are concerned with wall-detailing considerations.
Reinforcement The minimum web reinforcement ratio should be 0.0025,
which will satisfy requirements for both in- and out-of-plane loads. Reinforcement should be placed in at least two curtains and fully developed at the wall
boundaries. Section 21.9 of ACI 318-08, “Special Structural Walls,” contains
additional useful design recommendations and commentary.
Connected Elements Walls that may be subjected directly to transverse blast
pressure and that also provide gravity support for elements in higher stories
should be constructed with columns at each end and a beam spanning between
the columns. The purpose of this is to provide a gravity load path in the event
that the wall is compromised by the transverse blast load. The beams, columns,
and joints should be designed per previous sections of this chapter.
REFERENCES
This chapter is adapted in large part from information in the PCA’s Blast Resistant Design
Guide for Reinforced Concrete Structures (Portland Cement Association 2009), with
the permission of the Portland Cement Association.
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
(ACI 318-08) and Commentary (ACI 318R-08). Farmington Hills, MI: American
Concrete Institute.
REFERENCES
381
American Welding Society. 2005. Structural Welding Code—Reinforcing Steel
(ANSI/AWS D1.4). Miami, FL: American Welding Society.
ASTM. 2006. Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement (A706). West Conshohocken, PA: ASTM International.
. 2008. Specification for Deformed and Plain Billet-Steel Bars for Concrete Reinforcement (A615). West Conshohocken, PA: ASTM International.
Bilow, D. and M. Kamara. 2007. Capacity of joints to resist impact loads in concrete
moment-resisting frame buildings. Presented at the Structural Engineering Institute
Structural Congress, Long Beach, CA, May 16–19, 2007.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5-1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2000. Recommended Seismic Design Criteria for New Steel Moment-Frame Buildings (FEMA 350). Washington, DC: Federal
Emergency Management Agency, Department of Homeland Security.
Hayes, J. et al. 2005. Can strengthening for earthquake improve blast and progressive
collapse resistance? Journal of Structural Engineering 131 (8): 1157–1177.
Portland Cement Association. 2009. Blast Resistant Design Guide for Reinforced Concrete Structures. Skokie, IL: Portland Cement Association.
Unified Facilities Criteria Program. 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
15
Blast-Resistant Design Concepts
and Member Detailing: Steel
Charles Carter
15.1 GENERAL
This chapter provides guidance on the design of steel structures for blast loads.
It is adapted from information in two references: “Defensive Design” (Shipe
and Carter 2003) and Defensive Design of Structural Steel Buildings (Gilsanz
et al. 2009), with the permission of the American Institute of Steel Construction
(AISC).
Considerations for defensive design usually fall into one of three general
categories: typical building designs, prescriptive building designs, and
performance-based building designs. Some designs combine prescriptive and
performance-based approaches.
15.1.1 Typical Building Designs
The majority of buildings receive no special treatment other than judicious attention to redundant configurations and robust connection designs so that the structural (and nonstructural) components are tied together effectively. The typical
details used in steel buildings inherently provide for redundancy and robustness,
with the capability for load redistribution through alternative load paths. This
fact has been demonstrated repeatedly when steel buildings have been subjected
to abnormal loadings.
In such cases, the following ideas can be beneficial:
r Configure the building’s lateral systems to provide multiple load paths from
the roof to the foundations. Multiple lines of lateral framing distributed
throughout the building are generally better than fewer isolated systems.
r Provide horizontal floor and roof diaphragms to tie the gravity and lateral
framing systems together.
r Minimize framing irregularities in both horizontal and vertical framing when possible. Horizontal and vertical offsets with copes and/or
383
384
r
r
r
r
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
eccentricities can reduce the available strength at member ends—or require
extensive reinforcement to maintain that strength.
Use multiples of the same shape, rather than changing girder and column
sizes. The additional strength in girders and columns that are heavier for
convenience could cost less or be free. The use of a smaller number of
different shapes in the building means a labor savings in fabrication and
erection, and often more than offsets the cost of the additional steel weight,
which is only around 25% of the cost of the structure once it is fabricated
and erected.
Remember that serviceability limit states indirectly add significant structural redundancy to steel framing. Usually, beams and girders are sized for
deflection or floor-vibration criteria, and girders and columns are commonly
sized for drift control. As a result, these elements have significant reserve
strength.
Use typical shear, moment, and/or bracing connections judiciously. Reserve
strength is gained at low cost if connection details are clean. It costs little
to fill the web of a girder with bolts using a single-plate or double-angle
connection.
Recognize other sources of redundancy and robustness inherent in steel
buildings, including: the common overstrength in the steel materials and
connecting elements, membrane action in the floor and roof diaphragms,
and the strength and stiffness contributions of nonstructural components.
With little—and sometimes no—modification, steel framing provides redundancy and robustness.
15.1.2 Prescriptive Building Designs
Some buildings receive special treatment through the application of prescriptive
criteria for design that go beyond those in the basic building code. While there are
a variety of prescribed criteria, such as the removal of a single column, there is
usually no attempt to characterize the exact effects of the blast. Instead, the goal
could be to reduce the probability of progressive collapse in areas not directly
affected by the blast.
In such cases, several strategies can be employed:
r Use a perimeter moment frame. This can result in a system that is significantly robust. As an extreme example, the perimeter moment frame
in the tube structure of each of the World Trade Center towers spanned
a hole about 140 ft wide before succumbing to the combined effects of
structural damage and fire, each likely at their lifetime maximum values.
More likely, the prescribed criterion will be the removal of a single column. In some cases, the framing will have enough redundancy to accept
column removal without modification. If not, the column spacing can be
GENERAL
385
reduced or the framing hardened by increasing size or switching to
composite construction.
r Use a strong story or floor. This solution can be a truss system with diagonals or a Vierendeel truss system, incorporated into a single story or multiple stories in the building. Additionally, a single, strong floor with heavier
framing and moment connections throughout could carry or hang at least
10 floors. The World Trade Center towers also demonstrate this solution.
The damaged core columns in each tower to some extent hung from the
hat trusses, which created strong-story framing at the top of each tower.
The link between the core gravity columns and the perimeter tube framing carried more than 10 stories in one tower and more than 25 stories in
the other. Nonetheless, exercise caution when considering hat-truss framing to create a strong story. Unless specifically designed for progressivecollapse resistance, hat trusses normally reduce the level of reserve strength
and redundancy because of the efficiency they allow in the structural
system.
r Consider other innovative solutions. One particularly innovative solution
was used in the U.S. Federal Courthouse in Seattle, WA. The building
has a steel-framed composite core and gravity steel framing around it. The
perimeter has steel cables banding it to prevent progressive collapse should
a column be lost. The result is one of the most open and inviting blasthardened buildings built to date.
There are other potential prescriptive criteria, but structural solutions usually
flow from criteria like those described above. Most often, there is no attempt
to rigorously assess the actual blast effects and the emphasis is on arresting the
effects of the blast.
15.1.3 Performance-Based Building Designs
Performance-based criteria are used for a small number of buildings, normally
ones that are government-related, high-risk, or high-profile. Performance criteria vary, but generally require that the building withstand the effects of the blast,
protect the occupants of the building, and/or maintain a defined level of operability. The nature and characteristics of the threats are identified realistically and
modeled in the design. Note that the performance criteria affect more than the
structural frame, and often require nonstructural elements, such as blast-resistant
windows, special site layouts, and site perimeter protection.
Performance-based criteria are used for a comparatively small number of
buildings, where the nature and characteristics of the threat are realistically identified and modeled in the design. When the threat, building characteristics, and
performance criteria are known, the solution is a matter of design for the dynamic loading of the blast, including mitigation of damage and the associated
progressive collapse potential. The following sections illustrate this approach.
386
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
15.2 BLAST EFFECTS ON STRUCTURAL STEEL AND COMPOSITE
STRUCTURES
Steel’s response to the characteristics of a blast affects the loading for several
reasons, discussed in the following subsections.
15.2.1 Member Ductility
Structural steel has a linear stress-strain relationship up to the yield stress, but
then can undergo extreme elongation without an increase in stress—about 10 to
15 times the amount needed to reach yield. Stress then increases in the “strain
hardening” range until a total elongation of about 20 percent to 30 percent is
reached. This response has benefits beyond routine design-level forces for resisting the effects of a blast. The ductility ratio m, defined as the maximum deflection
over the elastic deflection, commonly is used to account for this effect.
15.2.2 Connection Ductility
Steel connections can be extremely ductile. As in seismic design practice, connections can be configured so that yielding limit states will control over fracture
limit states. However, there is still significant ductility in many limit states that
involve fracture. Block shear rupture, for example, is a fracture limit-state accompanied by significant deformation prior to the actual failure.
Additionally, some limit states have “failures” that, in blast applications, can
be considered benign or even beneficial. For example, there is little structural
consequence to connection slip or bearing deformations at bolt holes due to blast
effects. Moreover, these phenomena consume energy when they occur, progressively reducing the blast effect through the structure.
15.2.3 Overstrength
Structural steel usually is stronger than the specified minimum strength. For
example, ASTM A992, Standard Specification for Structural Steel Shapes
(ASTM 2006) has a specified minimum yield strength Fy = 50 ksi. From the
American Institute of Steel Construction (AISC) Seismic Provisions for Structural Steel Buildings (American Institute of Steel Construction 2005a), it has an
expected yield strength Fye of 55 ksi. This supports the practice endorsed by the
Department of Defense Explosives Safety Board, which allows the use of an average yield strength increase factor (generally a 10 percent increase in the yield
strength) for blast design.
15.2.4 Beneficial Strain-Rate Effects
Structural steel benefits from an increase in apparent strength when the rate of
loading is rapid. The yield point increases substantially by a factor that is called
BLAST EFFECTS ON STRUCTURAL STEEL AND COMPOSITE STRUCTURES
387
the Dynamic Increase Factor for yield stress. The ultimate tensile stress also
increases, but not as greatly as the yield stress. The total elongation at failure
typically remains unchanged or decreases slightly because of the increased strain
rate. The modulus of elasticity is unaffected by the strain rate.
15.2.5 Beneficial Effects of Composite Construction
The use of composite construction can have benefits for blast-design applications
due to the mass effect of composite systems with steel elements (490 lb/ft3 )
and concrete elements (150 lb/ft3 ). The inelastic action in a composite system
generally will limit deflections and local deformations, and partially mitigate
rebound effects through the damping effect of concrete cracking.
15.2.6 Perimeter Column Design
At the building perimeter, the columns will be loaded by the blast through the
facade. A critical design decision is the selection of an appropriate tributary area
for the column, which must be based upon the expected performance of the facade in the blast. Will the facade and its connection to the structure be such that
the full blast load will be delivered to the framing, or will the facade components
be shattered in the blast before the load can be delivered through them?
The tributary area is then used with the maximum total dynamic pressure and
the dynamic load factor to find an equivalent static load on the column. Either a
rigorous beam-column design approach or the simplified approach suggested in
UFC 3-340-02 (Unified Facilities Criteria Program 2008) can be used:
r Support conditions for the column are taken as if the column were a simplespan beam.
r The moment perpendicular to the plane of the facade is taken as wl2 /8 based
upon the blast loading.
r The moment in the plane of the facade is a function of the magnitude of the
strong-axis moment and the angle of incidence of the blast to the column.
r The axial load applied concurrently with the blast load is based upon the
full dead load and one-quarter of the live load.
Based upon these loadings, UFC 3-340-02 recommends checking limit states
like flange-local buckling, web-local buckling, shear yield, and interaction between flexural and axial forces.
15.2.7 Perimeter Girder Design
The consequences of a girder failure normally are not as high as the consequences of a column failure. When designing the girder to the same performance
level as the columns, use the foregoing approach to column design for the girder
388
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
design, without the axial load. Otherwise, a less restrictive approach can be used,
permitting inelastic deformations. It is important to consider the differing support conditions between the top and bottom flanges of the girder. Often this can
be mitigated by orienting the infill beams to provide restraint to the girder bottom flange along the span. Also, the girder end connections often will be best
configured as moment connections.
15.2.8 Slab Design
The blast pressure could subject the slab to significant uplift, depending upon
the expected performance of the facade in the blast. As with girder design, the
consequences of a local slab failure normally are not as high as the consequences
of a column failure, so it could be sufficient to utilize the typical reinforcing steel
in the upper portion of the slab. If a higher performance objective is established,
the blast pressure can be applied to the slab and the reinforcement selected to be
appropriate for the loading.
15.3 ANALYSIS AND DESIGN OF STRUCTURAL MEMBERS
Failure modes determine the mechanism of collapse for the structural element.
Deformation criteria classify the response of the element. The primary failure
modes used to design structural elements are breaching, tension-compression,
bending, shear, and axial-bending interaction.
Member failure is defined through support rotation and ductility ratios for use
with single-degree-of-freedom (SDOF) and simplified multi-degree-of-freedom
(MDOF) systems. Strain-based failure criteria may also be justified if strains are
calculated using finite element methods that characterize the material properties, the details of construction, and many of the possible failure mechanisms.
Ductility is the ratio between the maximum deflection and the maximum elastic
deflection. This parameter is smaller than one (1) if the behavior is elastic, and
larger than one (1) if the behavior is plastic.
Energy dissipated along the load path is ignored, and it is assumed that the
elements are absorbing all of the energy, which is a conservative assumption.
The designer should give attention to the reactions and the connections along the
load path from one member to another one, as these will be required to transfer
the full energy of the system.
15.4 STEEL MATERIAL PROPERTIES FOR BLAST DESIGN
The properties used in conventional design of steel buildings are modified when
considering blast loadings to account for actual strength level and dynamic
effects.
STEEL MATERIAL PROPERTIES FOR BLAST DESIGN
389
15.4.1 Strength Increase Factor (SIF)
For steel grades with Fy = 50 ksi or less, the average yield stress in current production is approximately 10% larger than this minimum specified value. Therefore, for blast design the yield stresses should be multiplied by a Strength Increase Factor (SIF) of 1.10. For higher-strength grades, this average is smaller
than 5% larger; therefore, no factor is used on those grades. The tensile strength
Fu is not factored in any case.
15.4.2 Dynamic Increase Factor (DIF)
Steel mechanical properties vary with the time rate of strain. As compared with
the static values normally used in design, the following properties vary for dynamic loading:
r The yield point increases substantially.
r The ultimate tensile strength increases slightly.
r Modulus of elasticity does not vary, and the elongation at rupture either
remains constant or is slightly reduced.
The factor used to modify the static stress due to dynamic load is the Dynamic Increase Factor (DIF). These factors are defined in Design of Blast Resistant Buildings in Petrochemicals Facilities (American Society of Civil Engineers
1997) and UFC 3-340-02.
The values summarized in Table 15.1 are based upon an average strain ratio of
0.10 in./in./sec, which is characteristic of low-pressure explosions. Higher values
of strain ratio will result in larger values of DIF. UFC 3-340-02 provides values
of DIF for different average strain rates.
Common grades in use today, such as ASTM A992, A500, and A53 (ASTM
2006, ASTM 2007b, ASTM 2007a) are not addressed in the aforementioned
Table 15.1 Dynamic Increase Factors (DIF) for Structural Steel
DIF (Low Pressure)
Yield Stress
Material
Bending/Shear
Tension/Compression
Ultimate Stress
A36
1.29
1.19
1.10
A588
1.19
1.12
1.05
A514
1.09
1.05
1.00
A446
1.10
1.10
1.00
390
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
references. All are normal construction grades, similar to ASTM A36 (ASTM
2008), and suitable for use with the same DIF values.
15.4.3 Dynamic Design Stress
Based on the expected ductility ratio and/or the damage allowed in the structural
element, different dynamic yield points are defined by UFC 3-340-02. If the
ductility ratio is smaller than or equal to 10, the dynamic design stress is:
f ds = f dy = SIF · DIF · f y
(15.1)
If the ductility ratio is greater than 10, the dynamic design stress is:
f ds = f dy +
f du − f dy
4
(15.2)
where
f du = DIF · f u
(15.3)
The dynamic design stress for shear is
f dv = 0.55 f ds
(15.4)
For typical design, unless governed by more detailed requirements such as
Design of Blast Resistant Buildings in Petrochemical Facilities (ASCE 1997),
UFC 3-340-02, or other established criteria, a simplified value may be used as
follows:
f ds = 1.3 f y
(15.5)
f du = 1.05 f u
(15.6)
15.5 DESIGN CRITERIA FOR BLAST DESIGN
15.5.1 General
Due to the nature of blast loading, plastic design is required, and is measured by
support rotations and ductility. Different codes can be used to define the strength
of the elements, but no safety factor should be used in those calculations. Local
and global stability should be addressed in those elements where plasticity must
be achieved and ductility criteria must be used.
Design and failure criteria are based on the results of explosive tests conducted
by the U.S. government and reported in publications such as the Tri-Service
DESIGN CRITERIA FOR BLAST DESIGN
391
Manual by the Department of Defense, and the Protective Design Manual by
the U.S. Air Force.
The designer should determine, at the beginning of the project, the acceptance
criteria to be used based on the level of protection desired for the building. Most
projects have predefined criteria that must be used.
15.5.2 Load Combinations
In the absence of other governing criteria, the following load combination should
be used:
1.0D + 0.25L + 1.0B
(15.7)
where D is the dead load
L is the live load
B is the blast load
15.5.3 Resistance Factor and Factor of Safety
For Load and Resistance Factor Design (LRFD), the resistance factor φ is not
used in blast design. For Allowable Stress Design (ASD), the safety factor
is not used in blast design. Thus, the strength is simply taken as the nominal
strength, without reduction.
15.5.4 Local Buckling
Blast design is based on the ultimate strength of the elements and ductility of the
system. Structural members subject to blast loads should be capable of undergoing plastic deformation. To allow hinges to form in the elements, sections must
be flexurally compact in accordance with criteria in Chapter B of the AISC Specification for Structural Steel Buildings (American Institute of Steel Construction
2005b), unless the effects of local buckling can be tolerated in the response. The
dynamic design stress should be used in the calculation of the limiting slenderness to establish the local buckling criteria.
15.5.5 Lateral-Torsional Buckling
Flexural members should be braced at an interval that is small enough to permit
plastic hinge formation. Essentially, this removes lateral-torsional buckling as
the controlling limit state in beams.
15.5.6 Deformation Criteria
Support rotation is defined as the tangent angle at the support formed by the
maximum beam deflection. In the plastic range, this value, neglecting the elastic
deformation, can be related to the plastic hinge rotation. Note that if the hinge is
392
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.1 Hinge Rotation versus Connection Rotation
not formed at the center of the beam, the support rotations are different at both
sides, and the maximum rotation should be considered as shown in Figure 15.1.
Ductility, µ, as used in blast design, is defined as the ratio between the maximum deflection (m ) and the elastic deflection E ).
µ=
m
E
(15.8)
ASCE’s Design of Blast Resistant Buildings in Petrochemical Facilities
(American Society of Civil Engineers 1997) classifies the deformation range in
three different damaged stages: low, medium, and high response as a function
of the damage in the building. UFC 3-340-02 also classifies the response as a
function of the protection provided by the structural elements. This section is
based upon a low level of protection that implies a high response as the intent is
to provide a design that avoids imminent collapse.
For different structural elements, this guide follows the response criteria as
shown in Table 15.2. The rotation criteria in Table 15.2 refer to support rotation.
The criteria are defined for the behavior of a single element. These criteria apply
to the design of the members, and the connection between the members may
have different criteria.
Table 15.2 Response Criteria for Structural Steel
for a Low Level of Protection
Element
Response Ratio (Ductility) Rotation
Open-web steel joists1
µ<2
θ < 3◦
Steel beams
µ < 20
θ < 10◦
Steel columns
µ<5
θ < 2◦
1
Response ratio (ductility) controlled by downward loading, and
rotation controlled by upward loading
DESIGN CRITERIA FOR BLAST DESIGN
393
As an example, assume a beam with a 30-ft span that starts to yield at midspan
when the deflection is 4 in. A support rotation of 10◦ implies a deflection of 15 ×
tan(10◦ ) = 31 in. A ductility of 1 implies a deflection of 4 in. Thus, a ductility of
20, again taken from Table 15.2, would require a deflection of 80 in. Obviously,
this level of ductility is not achievable, and the support rotation controls.
Other more restrictive criteria exist, such as Corps of Engineers Program Design and Construction (COE PDC) criteria, and are usually used in commercial
and governmental buildings.
Ductility limits are linked to a given mode of response. A flexural ductility is
different from shear or tension. For steel elements not shown in Table 15.2, other
sources can be consulted, including UFC 3-340-02 and Design of Blast Resistant Buildings in Petrochemicals Facilities. Although it is intended for seismic
design, FEMA 356 (Federal Emergency Management Agency 2000) provides
useful information.
UFC 3-340-02 defines the blast criteria and the structural properties used in
Figure 15.2.
15.5.7 Detailing for Specific Failure Modes:
Failure modes include breaching, tension, compression, shear, flexure, and combined loading.
r Breaching: Blast loads in contact with or in very close proximity to an element. May cause failure before a more typical structural response like flexure can occur, due to the high pressure produced by the explosion. Generally, if the scaled distance is below two, it is possible for the element to
breach before the overall response of the structural element starts.
For very close charges, in addition to the air blast, temperature and the
shock wave are important. An explosion in direct contact with a floor or wall
will interact directly with it and induce a shock wave inside it. The speed
and magnitude of the shock waves can cause the floor or wall to crack internally. For example, when the shock wave reaches the opposite face of a concrete wall, a section of concrete may spall or separate from the wall because
the energy of the shock wave exceeds the tensile strength of the material.
This behavior can be estimated using information in Figure 15.3, from
the Protective Construction Design Manual, ESL TR 87-57 (Air Force Engineering and Services Center 1989), based upon the thickness (T, ft), the
standoff distance (R, ft), and the charge (W, equivalent lbs TNT). Depending on the performance requirements, the slab may be designed to allow a
local breach in the bay closest to the blast, yet withstand the blast in adjacent bays. Typically, the standoff and charge weight are specified, given the
threat analysis. Other breaching curves can be found in ESL-TR-87-57 or
UFC 3-340-01 (Unified Facilities Criteria Program 2002).
r Tension: According to AISC’s Specification for Structural Steel Buildings,
the maximum strength for elements under tension is based on yielding of
394
Figure 15.2 UFC 3-340-02 Criteria
DESIGN CRITERIA FOR BLAST DESIGN
395
Figure 15.3 Protective Construction Manual Breach Chart
the member and is defined by
Pu = φ f y A g
(15.9)
where A g = the gross area of the element
f y = the material yield stress
φ = the resistance factor, 0.90
For blast loading, φ is not used, and f y = f ds is the dynamic design stress
defined in Equations 15.1 and 15.2.
Axial hinge properties for these kinds of elements are elastic-plastic,
based on the ultimate load defined above. If necessary, for computational
stability, a small strain hardening slope, such as 0.1%, can be used in the
plastic area.
r Compression: The strength of compression elements, based on the AISC
Specification for Structural Steel Buildings, is governed by the following
equation:
Pu = φ f cr A g
where A g = the gross area of the element
f cr = the critical stress
φ = 0.90
(15.10)
396
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For blast loading, φ is not used, and f cr = the critical stress by the appropriate AISC equation substituting f ds for f y . For a tension-compression
hinge, the recommendations of FEMA 356 can be used, if no other proper
criterion is defined.
r Shear: For I-shapes it is assumed that moment is carried primarily by the
flanges, while shear is carried primarily by the web. Thus, moment-shear
interaction is neglected. The yield strength of elements in shear, based on
the AISC Specification, is governed by the following equation:
V p = φ f v Aw
(15.11)
where Aw = the area of the web
f v = 0.6 f y
φ = 1.00 for most W-shapes
For blast loading, φ is not used, Aw is the area of the web, and f v = f dv is
the dynamic yield stress; f v = 0.55 f y in UFC 3-340-02 (Unified Facilities
Criteria Program 2008). Note that connection strength is, in most of the
cases, more critical than the shear strength of the beam.
r Flexure: Based on plastic response allowed for blast design, a plastic hinge
will be allowed to form in the structural element. The assumption that the
plastic hinge is concentrated at a section will be taken as adequate for practical purposes, even though deflection values may not be accurate.
The definition of plastic moment is a function of the ductility ratio expected in the structural element. Based on Figure 15.7 of UFC 3-340-02, if
the ductility ratio is smaller than 3, the full plastic moment cannot be developed by the section. Thus, an average between the elastic and plastic section
modulus is used:
M p = f ds
S+Z
2
(15.12)
If the ductility ratio expected is larger than 3, the full plastic moment can
be developed. Thus,
M p = f ds Z
(15.13)
For hand calculations, the behavior of the section is assumed elasticperfectly plastic. For hand calculations in this guide, the plastic moment
strength of the section is given by Equation 15.13. To avoid computational instabilities, the hinge properties used for computer calculations
should include plastic hardening, where the yield moment is defined by
Equation 15.12 and the ultimate moment is defined by Equation 15.13 as
EXAMPLES
397
recommended in UFC 3-340-02. A criterion for lateral bracing is also given
in UFC 3-340-02 although not discussed here.
r Combined Forces: Provisions for combined axial and bending effects are
presented in the AISC Specification. The strengths and resistance factors
defined above are used in the appropriate interaction equations.
15.6 EXAMPLES
15.6.1 Example 1—Determining Capacities
The strength of the elements determined in these examples is based on UFC
3-340-02. For element design, it is assumed that there is no energy dissipation
along the load path. A more accurate and less conservative procedure is to use
the dynamic reactions of one member on another along the load path.
When using member reactions to load following members, natural frequencies
should be compared. If the natural frequencies are close, a simultaneous solution
is required to account for the interaction between the two members. If the period
of the primary element (i.e., beam) is at least twice the period of the secondary
element (i.e., girder), they can be treated as individual SDOF structures on unyielding supports. If not, an MDOF solution of the same system should be used.
Material Properties All calculations for this example are done for steel with
the following properties:
r Fy = 50 ksi
r Fu = 70 ksi
For a ductility ratio µ ≤ 10, the dynamic design stress, defined in Equation
15.5, is:
f ds = f dy = 1.3Fy = 1.3 × 50 = 65 ksi
For a ductility ratio µ > 10, the dynamic design stress, defined in Equation
15.2, is:
f du − f dy
1.05 × 70 − 65
= 65 +
= 67 ksi
f ds = f dy +
4
4
where: f du = 1.05fu from Equation 15.6.
The dynamic design shear stress, defined in Equation 15.4, for a ductility ratio
µ ≤ 10 is:
f dv = 0.55 × f ds = 0.55 × 65 = 36 ksi
398
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.4 Tension Hinge Property
Tension-Only Rod Brace A tension rod with 3/4 -in. diameter is used. Because
the rod has upset ends, we can use the full cross section area. The tension strength
is obtained as:
Tmax = A f ds = π ×
3/4
2
2
× 65 = 28.7 kips
The hinge property for this rod follows the elastic-plastic curve shown in
Figure 15.4. A slight slope is included in the plastic region for computational
convergence purposes.
Given a maximum horizontal displacement of 4.2 in., the maximum rotation
at the base of the column for this displacement is:
α = atan
H
4.2
= atan
15 × 12
= 1.34 deg < 2 deg
For a small-displacement formulation, the elongation in the rod can be determined. First, the angle of the diagonal is determined:
θ = atan
H
L
15
= atan
35
= 23.2 deg
EXAMPLES
399
Then, the elongation, assuming the top of the column moves horizontally:
L = cos θ = 4.2 × cos 23.2 = 3.86 in
Thus, the elastic deformation of this rod is:
L el =
f ds L
65 × 38 × 12
=
= 1.02 in
E
29000
Therefore, the ductility demand in the rod is:
µ=
3.86
= 3.78
1.02
Diagonal Brace (in Tension and Compression) An ASTM A500 Grade C
HSS6 × 6 × 3/4 is used. This section has an area of 5.24 in.2 and a radius of
gyration of 2.34 in. The brace length is 17 ft.
The slenderness of this element is:
1 × 17ft × 12in. ft
2π 2 E
KL
=
= 88 < Cc =
= 94
λ=
r
2.34 in
f ds
The buckling stress is obtained from UFC 3-340-02 as:
⎛
2 ⎞
Kl
⎟
⎜
r
⎟
⎜
⎟ f ds
⎜1 −
⎝
2Cc2 ⎠
Fa =
5 3
+
3 8
Kl
r
Cc
−
1
8
Kl 3
r
Cc3
(15.14)
Substituting values, the nominal buckling stress is:
882
1−
× 65
2 × 942
Fa =
= 19.2 ksi
5 3 88 1 883
+ ×
− ×
3 8 94 8 943
The denominator in the above equation (which equals 1.92), is the factor of
safety. Hence, the compression capacity without the factor of safety is:
Pn = 1.92AFa = 1.67 × 5.24 × 19.2 = 193 kips
The tension capacity for this element is:
Pu = A f ds = 5.24 × 65 = 341 kips
400
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.5 Tension-Compression Hinge Properties
Using information from FEMA 356 (Federal Emergency Management
Agency 2000), a tension-compression hinge can be developed, as shown in Figure 15.5. This asymmetric curve includes the buckling ductility. In the tension
zone, the hardening slope is modified to account for the dynamic yield stress
for the yield axial force and the dynamic ultimate stress for the ultimate axial
force. In the compression zone, a hardening slope of 0.1% is introduced to avoid
computational instabilities.
Column Assume the column is a W12×53 with an effective length KL = 15 ft.
The section properties are:
A = 15.6 in.2
rmin = r y = 2.48 in.
For this column:
Cc =
2π 2 E
=
f ds
2 × π 2 × 29000
= 94
65
EXAMPLES
401
The maximum slenderness of this element is:
15 × 12
KL
= 72.6 < 94
=
ry
2.48
The buckling stress is:
72.62
1−
× 65
2 × 942
Fa =
= 24 ksi
5 3 72.6 1 72.63
+ ×
− ×
3 8
94
8
943
And the compression capacity without the factor of safety is:
Pu = 1.90AFa = 1.90 × 15.6 × 24 = 711 kips
The tension capacity of this element is:
Tu = A f ds = 15.6 × 65 = 1014 kips
Beam Assume the beam is a W12×35, 24-ft long, and it has chevron braces
connected to it at midspan. This shape has:
I = 285 in4
S = 45.6 in3
Z = 51.2 in3
It is modeled as a simply supported beam with a concentrated load at midspan,
where the braces meet.
For ductility ratios smaller than 3, the elastic-plastic bending moment is given
by
M pl =
S+Z
45.6 + 51.2 65
f ds =
×
= 262 kip − ft
2
2
12
The elastic deflection for this moment due to the concentrated load at the
midspan is
=
M pl L 2
262 × 12 × (24 × 12)2
=
= 2.63 in.
12EI
12 × 29,000 × 285
This deflection gives an elastic rotation:
θ = arctan
L 2
2.63
= arctan
24 × 12 2
= 1.05 deg
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Moment
402
Mult = Z × fds (µ > 10)
Mpl =
S+Z
fds
2
Rotation
2θult
2θ
Figure 15.6 Beam Hinge Moment-Rotation
Note that the hinge rotation is twice this support rotation. These parameters
define an elastic-perfectly plastic moment-rotation curve. Since many programs
have convergence problems with a perfectly plastic zone, a sloped line should be
introduced in this plastic region. This slope is obtained from Figure 15.6 for ductility greater than 10. Therefore, the ultimate bending moment is computed as:
Mult = Z × f ds (µ > 10) =
51.2 × 67
= 286 kips − ft
12
The maximum rotation from UFC 3-340-02 is:
θult = min (10 deg, 20 × θ) = min (10 deg, 26 deg) = 10 deg
Again, note that hinge rotation is twice this support rotation.
15.6.2 Example 2—Design and Analysis for Blast Loads on Members
Figures 15.7 and 15.8 illustrate the method used in this example for a simply
supported beam with load and mass uniformly distributed.
The same transformation as shown in Figure 15.7 can be done by multiplying
only the mass by the load-mass factor.
MSDOF = M K L M = M
KM
KL
The KLM approach is simpler because it only uses one transformation factor.
This example uses the load factor KL and the mass factor KM because they have
a more physical interpretation.
EXAMPLES
403
EQUIVALENT SDoF PROCEDURE
Continuous System
Discrete System
FSDoF(t)
F(t ) = p(t ) × L
MSDoF
M=m×L
KSDoF
L
Mass = M = m × L
Equivalent Mass = MSDoF = M × KM
Load = F(t ) = p(t ) × L
Equivalent Load = FSDoF (t ) = F(t ) × KL
Elastic Stiffness = K =
384 × E × I
5 × L3
Equivalent Stiffness = KSDoF = K × KL
Plastic Stiffness = Kpl = 0
Yield Force = Ryield =
8 × Mp
L
Equivalent Yield Force = Ryield, SDoF = Ryield × KL
Figure 15.7 MDOF to SDOF Simplification
Wind Girt The wind girt shown in Figure 15.9 has been designed as an MC8
× 20 (ASTM A572 Grade 50; Fy = 50 ksi; Fu = 65 ksi) for wind effects with
a deflection limitation of L/260. The blast deflection criteria given in Table 15.2
show that the ductility should be less than 20, and support rotation should be less
than 10 degrees.
The duration of the blast is 6.2 milliseconds, and the peak load for the 79.5
psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 5 × 25 = 1431 kips
1000
Also given, the load factor K L = 0.50 and the mass factor K M = 0.33.
As a preliminary design, the support rotation criterion is used because it does
not imply the knowledge of the actual section used. Rigid-perfectly plastic behavior is assumed, and the element is sufficiently braced against lateral-torsional
buckling.
404
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
SDoF NON-LINEAR DYNAMIC SOLUTION SYSTEM
System to be solved
FSDoF(t)
MSDoF Structural Period = T
KSDoF
IF
td
< 1
T 10
IMPULSIVE SOLUTION
IMPULSE ENERGY = STRAIN ENERGY
RESULTS
SOLUTION
1 < td
TIME-HISTORY DYNAMIC
IF
10 T
Fpeak, SDoF
DYNAMIC ANALYSIS CAN BE SOLVED:
- INTEGRATION OF MOTION EQUATION
- CHARTS
- SDoF SOFTWARE
FSDoF (t)
td
t
Figure 15.8 SDOF Solution
Figure 15.9 Example Girt
SDoF TO MDoF
- TIME AND DISPLACEMENT
ARE INVARIANT
- FORCE AND REACTION
SHOULD BE TRANSFORMED
EXAMPLES
405
For this rotation criterion, the maximum displacement allowed is:
max =
25 × 12
× sin (10 deg) = 26 in.
2
The self weight of the facade is 40 psf. The total weight of the system is:
w=
40 × 25 × 5
= 5 kips
1000
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33. Therefore, the parameters used in the discrete system are:
Fpeak,SDOF = Fpeak K L = 1431 × 0.5 = 715 kips wSDOF
= wK M = 5 × 0.33 = 1.65 kips
The equivalent impulse in this single-degree-of-freedom system is:
ISDOF =
F peak,SDOF tblast
715 × 6.2 × 10−3
=
= 2.2 kips − sec
2
2
The total energy produced by the blast load in the single-degree-of-freedom
is:
WImpulse,SDOF =
2
ISDOF
=
2m SDOF
2.22
= 566 kips − in
1.65
2×
386
To limit the maximum displacement to comply with the support rotation criteria, the single-degree-of-freedom yield force should be:
R yield,SDOF =
WImpulse,SDOF
566
= 21.8 kips
=
max
26
The yield force for the continuous system is:
R yield =
21.8
= 43.6 kips
0.5
For this yield force, the maximum moment is:
M pl =
R yield L
43.6 × 25
=
= 136 kips − ft
8
8
406
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For a maximum tensile strength of 65 ksi, the plastic modulus should be
greater than:
Z min =
136 × 12
= 25 in3
65
Based on this preliminary design, there is no MC8 strong enough to support
the blast. There are several possible modifications to improve the behavior of the
system: Increase the excited mass, increase the strength-stiffness of the system,
or decrease the blast load by integrating a variable blast pressure that is a function
of the distance to the charge along the girt.
In this example we will increase the steel section to a W8 × 28 (ASTM A992;
Fy = 50 ksi; Fu = 65 ksi):
Z x = 27.2 in3
M pl = Z f ds =
27.2 × 65
= 147 kip − ft
12
Ix = 98 in4
el =
5 147 × 12 × (25 × 12)2
5 M pl L 2
=
= 5.82 in
48 E I
48
29,000 × 98
The strength parameters to use in the dynamic calculation are:
8M pl
8 × 147
=
= 47.0 kips for the SDOF
L
25
= R yield K L = 47.0 × 0.5 = 23.6 kips
R yield =
R yield,SDOF
R yield
47
= 8.08 kips in for the SDOF
=
el
5.82
= KK L = 8.08 × 0.5 = 4.04 kips in
K =
K SDOF
The mass and load do not change from the latest calculations.
The period of the system is:
T = 2π
m SDOF
=2×π ×
K SDOF
1.65
= 0.20 sec
386 × 4.04
The structural period (0.20 sec) is more than 10 times longer than the load
duration (0.0062 sec), hence the assumption of impulsive load is correct.
Figure 15.10 shows the displacement computed using SDOF software. The
maximum deflection is 28.8 in. > 26.0 in.; hence, the rotation criterion is not
met by a slight margin.
EXAMPLES
407
Figure 15.10 SDOF Displacement
The elastic displacement for this beam is 5.82 in.; therefore, the ductility ratio
for this system is:
µ=
28.83
= 5 < 20
5.82
The maximum reaction happens at the time of the maximum response. In an
impulse load, at the time the response is maximum, the load is zero. For a longer
blast duration, the reaction is a combination of the direct load and the capacity
of the resisting element. According to UFC 3-340-02, the maximum shear force
in plastic design can be obtained as the shear reaction that corresponds to the
elastic limit of the section:
V =
R yield
47.2
=
= 23.6 kips
2
2
The dynamic shear capacity is:
V p = f dv Aw = 36 × 2.30 = 82.8 kips > 23.6 kips
where f dv is the blast shear capacity.
Aw is the area of the web
408
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Hence, the section can support the shear. The connection should be designed
for this capacity.
Facade Column The column shown in Figure 15.11 has been designed as a
W12 × 53 (ASTM A992; Fy = 50 ksi; Fu = 65 ksi) for wind effects with a
deflection limitation of L/260. The properties of this section are:
A = 15.6 in.
Ix = 425 in .4
r y = 2.48 in.
Z x = 77.9 in.3
r x = 5.23 in.
As for the wind girt, the duration of the blast is 6.2 milliseconds, and the peak
load for the 79.5 psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 5 × 25 = 1431 kips
1000
Also given, the load factor K L = 0.50 and the mass factor K M = 0.33.
Columns are designed to remain elastic; therefore, the maximum bending capacity should not be reached. As a preliminary design, the column will be designed without axial compression. Combined bending compression formulation
will be used to check the adequacy of this element.
Figure 15.11 Example Column
EXAMPLES
409
As a preliminary and conservative assumption, the girt is assumed infinitely
rigid, and all the blast pressure is absorbed by the column.
The self weight of the facade is 40 psf. The total weight of the system is:
w=
40 × 25 × 15
= 15 kips
1000
The peak load for the 79.5 psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 15 × 25 = 4293 kips
1000
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and the stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33.
Therefore, the parameters used in the discrete system are:
F peak,SDOF = 0.5 × 4293 = 2146 kips
wSDOF = 0.33 × 15 = 4.95 kips
The equivalent impulse in this single-degree-of-freedom system is:
ISDOF =
F peak,SDOF tblast
2146 × 6.2 × 10−3
=
= 6.65 kips − sec
2
2
The total energy produced by the blast load in the single-degree-of-freedom is:
WImpulse,SDOF =
2
ISDOF
=
2 m SDOF
6.652
= 1724 kip − in.
4.95
2×
386
For this W12 × 53, assuming a uniformly distributed load, the following properties define the structural behavior:
M pl = Zfds =
el =
77.9 × 65
= 422 kip − ft
12
5 422 × 12 × (15 × 12)2
5 M pl L 2
=
= 1.38 in.
48 E I
48
29,000 × 425
410
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The parameters to use in the dynamic calculation are:
8M pl
8 × 422
=
= 225 kips, for the SDOF
L
15
= R yield K L = 225 × 0.5 = 112.5 kips
R yield =
R yield,SDOF
R yield
225
= 163 kips/in., for the SDOF
=
el
1.38
= 163 × 0.5 = 81.5 kips/in.
K =
KSDO F
The period of the system is:
T = 2π
m SDOF
=2×π ×
K SDOF
4.95
= 0.08 sec
386 × 81.5
The column period (0.08 sec) is more than 10 times the load duration (0.0062
sec); hence, the assumption of impulsive load can be assumed. The period of
the column is smaller than half of the beam period (0.08 sec < 0.20
= 0.1 sec);
2
hence, the system is uncoupled and can be modeled separately.
Assuming elastic behavior, the maximum elastic energy that can be adsorbed
by the SDOF system is given by the equation:
W S,el,max =
2
Ryield
,SDOF
2K SDOF
=
112.52
= 77.6 kips − in. 1724 kips − in.
2 × 81.5
This energy is smaller than the energy induced by the impulse; hence, this element will achieve plastic behavior. There is not an economical solution to solve
this problem because of the element’s inability to remain elastic for a directly
applied tributary blast load. But the preliminary assumption that based this design in the rigid behavior of the girt is highly conservative; the maximum load
that this element is carrying comes from the reaction in the girt.
Assuming this reaction is static, the system to solve is defined in Figure 15.12.
The maximum bending moment for this configuration is:
Mx = 2 × 23.6 ×
15
= 236 kips − ft
3
The axial load at the column, based on 30 psf roof dead load, is:
P = 30 ×
50 70
×
= 13.1 kips
2
4
In the following, the capacity properties of this element are calculated to be
used in the interaction axial-bending curves.
EXAMPLES
411
Figure 15.12 Equivalent Static Reaction System
For buckling about the weak axis, the column is assumed unbraced by the
girts. The buckling length is K y L = 1 × 15 = 15 ft; hence, the slenderness is:
Ky L
15 × 12
=
= 72.6
ry
2.48
For the buckling about the strong axis, the column is also unbraced. The buckling
length is K x L = 1 × 15 = 15 ft; hence, the slenderness is:
15 × 12
Kx L
= 34.4
=
rx
5.23
Also,
Cc =
2π 2 E
=
f ds
2 × π 2 × 29000
= 94
65
And the allowable compression stress is calculated by Equation 15-14:
72.62
1−
× 65
2 × 942
Fa =
= 24 ksi
5 3 72.6 1 72.63
+ ×
− ×
3 8
94
8
943
412
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The nominal compression capacity results:
Pu = 1.90AFa = 1.90 × 15.6 × 24 = 711 kips
The Euler stress in the x-axis is:
Fex =
12π 2 E
12 × π 2 × 29000
=
= 129 ksi
23λ2x
23 × 342
In this equation, 12/23 is the factor of safety for buckling. And the unfactored
Euler load in the x-axis results:
Pex =
23
23
AFex =
× 15.6 × 129 = 3857 kips
12
12
The tension capacity is:
Pp = A f ds = 15.6 × 65 = 1014 kips
The plastic moment was calculated before as:
M pl,x = Z x f ds =
77.9 × 65
= 422 kip − ft
12
This value was calculated assuming the member was completely braced, but
this beam is laterally unsupported because the girts cannot carry any axial load
after the blast. Hence, the unbraced flexural strength is:
√
√
L/r y f ds
15 × 12/2.48 × 65
M pl,x = 1.07 −
Mm,x = 1.07 −
3160
3160
×422 = 0.88 × 422 = 371 kip − ft
The first interaction curve that the system should satisfy is:
P
1 × 236
Cmx Mx
13.1
+
+
=
P
13.1
Pu
636
1−
Mmx
1−
× 371
Pex
3857
= 0.015 + 0.64 = 0.65 ≤ 1
where
Cmx = 1
All the other parameters were defined previously.
The second interaction curve to accomplish is:
Mx
P
236
13.1
= 0.56 ≤ 1 for
= 0.013 < 0.15
=
=
M pl,x
422
Pp
1014
EXAMPLES
413
Figure 15.13 Composite Beam Section
This calculation does not consider the reduced stiffness due to the interaction
of bending and compression. For higher compression loads and if the moment
magnifier is bigger than 10 percent, the analysis should consider second-order
effects.
Composite Beam In this example, the composite beam shown in Figure 15.13
will be designed for the roof blast load. The structure is a 25-ft span composite
beam, W14 × 22 with a 51/2 slab with a 3 metal deck. The concrete strength
is 3.0 ksi. For blast design, this strength is multiplied by 1.12. The steel used is
Grade 50, as in the previous examples.
The section properties based on the blast strength for the concrete and the
steel are:
I = 830 in.4
M pl = 395 kip − ft
For rebound purposes, the moment capacity of the steel, based on Z =
33.2 in3 , is:
M pl,steel = Z f dy =
33.2(65)
= 180 kip − ft
12
The load to be used for this example is shown in Figure 15.14.
Finally, assume that the existing dead load is 60 psf.
The total weight for the beam is:
w=
60 × 25 × 6
= 9 kips
1000
The peak load for the 12.2 psi peak blast pressure is:
Fpeak =
12.2 × 12 × 12
× 6 × 25 = 263 kips
1000
414
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.14 Roof Blast Load
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and the stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33. Therefore, the parameters used in the discrete system are:
Fpeak,SDOF = 0.5 × 263 = 131.5 kips
wSDOF = 0.33 × 9 = 2.97 kips
The maximum elastic deflection is:
el =
5 395 × 12 × (25 × 12)2
5 M pl L 2
=
= 1.85 in.
48 E I
48
29000 × 830
The parameters to use in the dynamic calculation are:
R yield =
8M pl
8 × 395
=
= 126 kips
L
25
EXAMPLES
415
For the design of this beam, the existing dead load is applied simultaneously
with the blast load and will reduce the beam capacity.
R yield,r educed = 126 − 9 = 117 kips. For the SDOF
Ryield,SDOF = 117 × 0.5 = 58.5 kips
R yield
126
= 68 kips in. For the SDOF
=
el
1.85
= 68 × 0.5 = 34 kips in.
K =
K SDOF
The period of the system is:
T = 2π
m SDOF
= 2π
K SDOF
wSDOF
= 2π
gK SDOF
2.97
= 0.09 sec
386 × 34
For this example, the beam period is less than 10 times the load
duration,0.09 sec < 10(0.0144) = 0.144 sec; hence, the impulse formulation
cannot be used.
For the rebound
R yield,steel =
8M pl,steel
8 × 180
=
= 57.6 kips
L
25
The elastic deflection for the noncomposite steel beam is:
el,steel =
40 M pl,steel L 2
40 180 × 12 × (25 × 12)2
= 3.50 in.
=
384 E Isteel
384
29000 × 199
The maximum negative force is:
RRebound,SDOF = 0.5 × 57.6 kips = 28.8 kips
Therefore, the elastic stiffness at the rebound is:
K Rebound,SDOF =
RRebound,SDOF
28.8
= 8.22 kips in.
=
el,steel
3.50
The period at the rebound is:
T = 2π
m SDOF
K Rebound,SDOF
=2×π ×
2.97
= 0.19 sec
386 × 8.22
416
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For the rebound, the beam period is greater than 10 times the load
duration,0.19 sec > 0.0144 sec; hence, the impulse formulation can be used for
the rebound response.
The parameters for the load used in the analysis were previously calculated:
wSDOF = 2.97 kips
Fpeak,SDOF = 131.5 kips
The load is applied follow the timing shown in Figure 15.14.
The yield and stiffness were obtained previously as:
Ryield,SDOF = 58.5 kips
K SDOF = 34 kips/in.
And the rebound properties are:
RRebound,SDOF = 28.8 kips
K Rebound,SDOF = 8.22 kips in
Using nonlinear SDOF software, the maximum displacement is (see
Figure 15.15):
SDOF = 1.84 in.
SDOF,Rebound = 3.61 in.
Figure 15.15 SDOF Displacement
EXAMPLES
417
The dead load deflection should be added to the solution obtained by SDOF.
dead =
5 9 × (25 × 12)3
5 W L3
=
= 0.13 in.
384 E I
384 29000 × 830
The maximum deflection results:
max = SDOF + dead = 1.84 + 0.13 = 1.97 in.
Hence, the ductility results:
µ=
1.97
max
=
= 1.06
el
1.85
For the rebound, the ductility is:
µ=
3.61
= 1.03
3.50
For the SDOF solution the element behaves as a composite member and
yields during the rebound. Figure 15.16 shows the force resultant from the SDOF
system.
Figure 15.16 SDOF Force
418
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The maximum shear force in plastic design can be obtained as the shear reaction that plastifies the section:
V =
R yield
126
=
= 63 kips
2
2
The dynamic shear capacity is:
V p = f dv Aw = 36 × 2.76 = 99.4 kips > 63 kips
where f dv is the blast shear capacity
Aw is the area of the web
Hence, the section can support the shear. The connection should be designed
for this capacity
15.7 DESIGN OF CONNECTIONS
Connections require as much attention as members, especially in blast design.
Failures of steel structures under overload can initiate at the connections of members, as opposed to the members themselves, for many reasons:
r The design of members is often controlled by considerations other than
strength, including deflection, vibration control, and architectural requirements. Such members are often substantially stronger than required to resist the design forces. Connections, however, typically can be designed
based only on considerations of strength. When this is the case, there may
be less overstrength or reserve capacity to match the overstrength of the
members.
r Connections may be controlled by limit states with varying levels of ductility. When less ductile limit states, such as tension rupture or block shear
rupture control, there is less capacity for redistribution of forces and the
mobilization of ductile behavior.
r Connections tend to be of limited size and, therefore, even when exhibiting
ductile behavior, they can only accommodate limited plastic deformation
before reaching their ultimate capacities
It is important in the design of blast-resistant structures, particularly those
expected to be loaded into the inelastic range of behavior, that sudden failure
modes be avoided so that the plastic response of the structure can be mobilized.
It is also important to remember that blast loading may result in load reversal,
REFERENCES
419
Table 15.3 Dynamic Increase Factors for Connection Design
Connection Material
Action
Yield Strength
Ultimate Strength
Plate, A36, A572
Bending/shear
1.3
1.1
Tension/compression/bearing
1.2
1.1
Plate A588
Bending/shear
1.2
1.05
Tension/compression/bearing
1.12
1.05
Bolts
Shear/tension
1.0
1.0
Weld metal
Same as base metals
sometimes in the form of rebound. Connections should be designed with equal
strength under load reversal, unless the following apply:
1. Member strength is limited in one direction of load application by considerations of buckling, as will occur for braces in compression and flexural
members with one flange braced and the other unbraced.
2. Nonlinear dynamic analysis is performed to determine the maximum connection loads in both positive and negative loading applications.
Connections should be designed to develop the full plastic capacity of the supported members, so that the plastic response of the structure can be mobilized in
resisting blast-induced stresses. It should be noted that the dynamic plastic capacity of an element, loaded briefly by impulsive loading, is often greater than the
static plastic capacity. Alternatively, when nonlinear dynamic analysis of structures and structural elements under blast loading is performed, it is permissible
to design the connections for the peak forces obtained from the analysis.
The specified yield and tensile strength of the connection material, including
bolts, welds, angles, and plates, may be increased in accordance with Table 15.3
to account for the Dynamic Increase Factor (DIF). Note that Strength Increase
Factors (SIF), discussed previously for members, are not used in the design of
connections as an additional means of ensuring that connections will be capable
of developing the strength of the member.
REFERENCES
Air Force Engineering and Services Center. 1989. Protective Construction Design Manual. (ESL-TR-87-57). Tyndall Air Force Base, FL: Air Force Engineering and Services
Laboratory.
American Institute of Steel Construction. 2005a. Seismic Provisions for Structural Steel
Buildings. Chicago, IL: American Institute of Steel Construction.
. 2005b. Specification for Structural Steel Buildings. Chicago, IL: American
Institute of Steel Construction.
420
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
ASTM. 2006. Standard Specification for Structural Steel Shapes (A992). West
Conshohocken, PA: ASTM International.
. 2007a. Standard Specification for Pipe, Steel, Black and Hot-Dipped, ZincCoated, Welded and Seamless (A53). West Conshohocken, PA: ASTM International.
. 2007b. Standard Specification for Cold-Formed Welded and Seamless
Carbon Steel Structural Tubing in Rounds and Shapes (A500). West Conshohocken,
PA: ASTM International.
. 2008. Standard Specification for Carbon Structural Steel (A36). West Conshohocken, PA: ASTM International.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5−1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2000. Prestandard and Commentary for the
Seismic Rehabilitation of Buildings (FEMA 356). Washington, DC: Federal Emergency Management Agency.
Gilsanz, Ramon et al. 2009. Defensive Design of Structural Steel Buildings (AISC Design
Guide No. 27), Chapters 6 and 7. Chicago, IL: American Institute of Steel Construction.
Shipe, James A. and Charles J. Carter. 2003. Defensive design. Modern Steel Construction 43 (11): 25–31.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures to Conventional Weapons Effects (FOUO) (UFC 3-340-01).
Washington, DC: U.S. Department of Defense, Unified Facilities Criteria
Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
16
Blast-Resistant Design
Concepts and Member
Detailing: Masonry
Shalva Marjanishvili
Prior to the twentieth century, unreinforced masonry was the primary building
material for both residential and commercial construction. As a result, unreinforced masonry structures constitute a large portion of existing buildings, many
of which have historical and architectural importance.
Beginning in the early 1980s with terrorist attacks in Lebanon and Kuwait,
high-profile and high-risk civilian federal buildings have been designed to resist
the effects of explosive attacks. With each major attack on U.S. interests, more
attention has been paid to this low probability/high consequence threat scenario.
As a result, recent work has been dedicated to the design and analysis of masonry
structures to resist air-blast loads.
When a wall is first subjected to an air-blast load, it experiences out-of-plane
flexure, producing tensile strains on the interior face of the wall and compressive
strains on the exterior. Once the maximum positive deflection associated with
the out-of-plane flexure has been attained, the wall will begin to vibrate due to
rebound forces created by negative blast pressures. The wall undergoes negative
deflection and curvature, causing tensile strains to develop on the exterior face
of the wall, and increasing shear stresses at wall supports.
Concrete masonry unit construction is primarily used for walls. These walls,
when properly designed, will provide economical resistance to relatively low
air-blast pressures. Limitations of masonry walls include development of large
deformations and rebound response.
Most masonry design codes advocate elastic design to specified stress levels, even for dynamic loads. This approach is uneconomical and potentially unsafe because of its inability to predict structural response beyond the allowable
stress limit. Masonry walls, when subjected to air-blast loads, are expected to undergo large post-elastic response, and therefore, they should be designed using
the strength design methods.
Below are two primary assumptions made when assessing flexural strength of
reinforced masonry:
421
422
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Assumption 1.
Singly reinforced masonry units rely on composite action between compression in concrete and tension in reinforcement.
a. Plain sections remain plain during bending. At large deflections, significant deviation from this assumption may occur. This can be corrected by
including both bending and shear deformations in the analysis.
b. There is a perfect bond between grout and reinforcement.
c. Concrete tension strength is neglected.
d. Concrete stress block is idealized as a rectangle.
e. Reinforcement stress/strain is idealized as elastic-perfectly plastic with appropriate strain-rate hardening effects.
f. Flexural strength is attained when either tension reinforcement or extreme
fibers of concrete reach their ultimate strains.
Assumption 2.
Under extreme loading such as air blast, it is expected that the compressive
strength of masonry will be exceeded. When this happens, the resistance is provided purely by compression and tension in the reinforcement. This approach requires that masonry be reinforced with two curtains of reinforcement. Therefore,
doubly reinforced masonry units, under this assumption, rely on compressive and
tension forces in the reinforcement only.
a. Plain sections remain plain during bending.
b. Reinforcement is assumed to be held in its original position, as if the masonry were undamaged.
c. Concrete tension and compression strengths are neglected.
d. Reinforcement stress/strain relationship is idealized as elastic-perfectly
plastic with appropriate strain hardening effects.
e. Maximum flexural strength is attained when either curtain of reinforcement yields either in tension or in compression.
Assumption 1 is appropriate only for level of protection (LOP) I, the superficial damage level. For other damage levels (LOP II–LOP IV), it is necessary to
use Assumption 2. It is worth noting that reinforced masonry will often behave
under Assumption 2.
When a masonry wall is subjected to relatively small air-blast pressures with
low performance requirements (LOP I), ACI 530-08, Section 3.3.5 (American
Concrete Institute 2008) can be used to estimate out-of-plane moment capacity
of the wall, with the following modifications:
1. φ strength reduction factors shall be taken as one (i.e., φ = 1.0).
2. Yield strength of reinforcement shall be taken as dynamic (i.e., f y = f dy ).
3. Compressive strength of masonry shall be taken as dynamic (i.e., f m =
).
f dm
GENERAL CONSIDERATIONS
423
UFC 3-340-02 (Unified Facilities Criteria 2008) suggests the following equation for estimating ultimate moment capacity under Assumption 2:
Mu = As f dy dc
(UFC 3-340-02, Equation 6.2)
where: As = area of joint reinforcement at one face
f dy = dynamic yield strength of reinforcement
dc = distance between centroids of compression and tension reinforcement
16.1 GENERAL CONSIDERATIONS
Reinforced masonry subjected to service design loads (excluding air-blast) can
be designed using both allowable and strength design methodologies. For airblast design, only strength design methodology should be used, since reinforced
masonry elements subjected to air-blast loads are likely to experience large nonlinear deformations with relatively large strain and ductility demands. Therefore,
it is necessary to employ a design methodology that can account for response
well beyond linear limit states. Working stress design, although much simpler
then strength design, only accounts for response below nonlinear range, and implies that all reinforced masonry should be designed linearly. Furthermore, airblast design requires that structural elements respond in a ductile manner, with
large nonlinear deformations. For these reasons, allowable stress design should
not be used for air-blast design of reinforced masonry.
Although unreinforced masonry should not be used for new construction,
there are many historic buildings with existing unreinforced masonry walls.
When existing unreinforced masonry walls cannot be replaced, the stability of
existing unreinforced masonry walls should be assessed. The stability assessment should assume that the wall is simply supported if flexible supports are
provided. It should be permissible to assume arching action in the masonry wall
if rigid supports are provided.
There will be cases when air-blast design requirements contradict or largely
exceed the conventional masonry design required by building codes. The design
engineer should verify that the final masonry design meets all design load requirements, including blast design requirements. If the blast design requirements
countermand or contradict other masonry design constraints, the owner and architect should be appraised of the issue by the design professional, and a decision
should be reached by the group in order to resolve the conflicting requirements.
The design engineer should direct the detailing on the design documents such
that all requirements of the blast-resistant design reworking and renovation of an
existing structure are met. The design engineer should verify that the design documents clearly detail any required repairs of the existing structure, modifications
to the existing structure, or further work to the existing structure that is necessary
to achieve the LOP required of the modified existing structure.
424
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
16.1.1 Masonry
Maximum masonry strength, f m, should not exceed 6,000 psi for any LOP in
new construction. This strength limitation is intended to preclude nonductile failure modes, such as shear and/or crushing.
Minimum masonry strength, f m, should not be less than 1,500 psi for any
LOP in new construction.
The design engineer should determine the material properties of existing masonry structural elements to ensure proper design to achieve the LOP required
for the existing masonry structure.
UFC 3-240-02 suggests the following equations to determine effective moment of inertia:
Ieff = Ia =
In + Icr
2
Icr = 0.005bdc3
(UFC 3-340-02, Equation 6-6)
(UFC 3-340-02, Equation 6-7)
While ACI 530-08 suggests that Ieff is approximately equal to one-half of
gross moment of inertia, it also suggests the following equation for a more accurate estimate of effective moment of inertia:
Mcr 3
Mcr 3
≤ In ≤ 0.5Ig
+ Icr 1 −
Ieff = In
Ma
Ma
(ACI 530-08, Section 3.5.3)
This handbook recommends the ACI approach when using Assumption 1, and
the UFC approach when using Assumption 2. This recommendation is based on
the understanding that the UFC document covers reinforced masonry response
due to extreme loading conditions, while ACI-350-08 is concerned mainly with
elastic response.
16.1.2 Reinforcement
All reinforcement should comply with ASTM A615 (Grade 60) or A706 for all
LOPs in new construction (ASTM 2006, ASTM 2008a).
It is desirable to use reinforcement with high ductility demand capability. Currently only ASTM A706 reinforcement is considered to have high ductility capacity. However, according to UFC 3-340-02, it is sufficient to use ASTM A615
(Grade 60).
The modulus of elasticity of reinforcement should be taken as required by
ACI 350-08.
At strain rates characteristic of air-blast response, reinforcing steel exhibits
a significant increase in yield strength above static test values. For 60-ksi
reinforcement, UFC 3-340-02 recommends a dynamic increase factor of 1.17,
GENERAL CONSIDERATIONS
425
that is:
f dy = 1.17 f y = 1.17 × 60 ksi = 70.2 ksi
Reinforcement splices should be limited to zones that remain elastic during
blast response. Furthermore, not more than 50% of splices should be located
along a single line. That is, splices should be staggered along the length or height
of the wall and away from potential yield lines.
Splices should develop full tensile capacity of reinforcement and should conform to ACI 350-08. It is possible to use mechanical splices; however, the use of
mechanical splices is limited by the geometric constraints of concrete masonry
units. Use of splices should be limited.
16.1.3 Mortar
Mortar is a mixture of cementitious materials, aggregates and water. All mortars
should comply with ASTM Standard C270 (ASTM 2000). Mortars with lime
should be avoided, as lime decreases the strength of mortar.
16.1.4 Grout
Grout provides bonding of longitudinal reinforcement and adds strength, rigidity, and mass to the wall system. All grouts are required to conform to ASTM
Standard C476 (ASTM 2008b). It is recommended that blast masonry walls be
fully grouted.
16.1.5 Construction Methods
Two commonly used types of masonry construction are running bond and stack
bond.
With running bond, masonry units are normally staggered by at least onequarter of the length of the unit. With typical 8-in. masonry units, the joints are
offset by at least 4 in., and since the length of the units is 16 in., it is common to
offset units by half the length of the unit.
With stack bond, joints in each successive course line up vertically. This type
of construction is not desirable in air-blast-resistant construction, since there is
little or no continuity across the head joints (the head joint is the joint that runs
up and down the wall). The only continuity across the head joints is provided by
longitudinal reinforcement and grout.
Conventional construction methods of reinforced masonry walls include running bond and stacked bond with the maximum spacing of vertical reinforcement
at 48 in. on center (o.c.). This implies that not every cell is grouted.
For air-blast-resistant design, it is recommended that every cell be grouted
and at least one reinforcing bar be placed in each concrete masonry unit. For
example, with 16-in. masonry units the maximum reinforcement spacing should
426
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
not exceed 16 in. o.c., and every cell should be grouted. Since concrete masonry
units (CMUs) are of modular construction with a modular dimension of 8 in.,
reinforcement must be placed in increments of 8 in.
When designing with reinforced masonry, dimensional constraints need to be
accounted for as follows:
1. Actual dimensions of masonry units are 3/8 in. less than the nominal dimensions.
2. Reinforcement should be placed at increments of 8 in. o.c. (i.e., 8 in.,
16 in.).
3. Since the largest CMU block has a length of 16 in., it is recommended that
at least one reinforcing bar be placed in each masonry unit, resulting in a
maximum spacing of 16 in. o.c.
4. Masonry unit wall thickness will dictate the minimum cover dimension.
It may be very difficult to achieve double curtain reinforcement in 4-in.
masonry units.
5. Double curtain reinforcement can be achieved with 8-in. masonry units by
staggering longitudinal reinforcement as shown in Figure 16.9
16.2 FAILURE MODES
Failure modes of masonry walls are categorized as either ductile or brittle.
Ductile failure is preferred, and is generally characterized by bending failure
caused by yielding of tensile reinforcement. Brittle failure modes include shear,
buckling, and crushing. Figures 16.1 though 16.3 are photos depicting failures of
masonry construction. Figure 16.1 shows infill masonry wall performance when
structural collapse is initiated. Figure 16.2 depicts infill masonry wall failure
without collapse of bearing elements. In this case, the masonry wall could
potentially become heavy debris thrown into the occupied rooms. Both Figures
16.1 and 16.2 show masonry wall performance under extremely high air-blast
loads, while Figure 16.3 shows reinforced masonry wall performance under
moderate to low air-blast pressures. Although the masonry wall shown in Figure
16.3 experienced some cracking and stucco spalling, it performed satisfactorily.
These figures illustrate how reinforced masonry walls could provide economical
protection under relatively moderate air-blast pressures.
Table 16.1 correlates qualitative damage description with the damage levels
superficial, moderate, heavy, and hazardous and their corresponding levels of
protection from IV though I.
The first step in air-blast design of masonry walls is to establish the desirable performance level in terms of acceptable damage levels, as shown in Table
16.1. Then, dynamic analysis is conducted to calculate damage indicators, as described under each column for corresponding damage levels. If it is determined
that the damage indicators are exceeded for a desired performance level, the
FAILURE MODES
Figure 16.1 Failure of Building with Masonry Infill Walls
Figure 16.2 Failure of Masonry Infill Wall
427
428
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.3 Crack Pattern on Masonry Infill Wall
wall must be redesigned and re-analyzed. Thus, air-blast design is an iterative
process.
To simplify air-blast analysis and design procedures, often a single-degree-offreedom (SDOF) analysis is conducted with a predetermined deflected shape of
the wall. The results of this SDOF analysis are usually obtained as ductilities and
support rotations. These calculated ductilities and rotations are then compared to
allowable values, to determine damage limits (see Chapter 3, section 3.5 of this
book for allowable ductilities and rotations for reinforced masonry).
Table 16.1 also lists performance indicators for unreinforced masonry. Although unreinforced masonry is not recommended in new construction, there are
many existing buildings with unreinforced masonry facades which may require
evaluation under air-blast loads.
16.2.1 Flexure
Flexure provides a ductile failure mode. This is the preferred failure mechanism
and occurs in properly designed and supported masonry. This can be achieved by
designing masonry for the balanced condition whereby reinforcement will yield
just before the masonry concrete begins crushing. Figure 16.4 depicts typical
deflected shape due to flexural response.
When air-blast pressures are relatively low, it may be possible to design a reinforced masonry wall with just one curtain of reinforcement located in the middle.
429
Limit State
Multiple
Multiple
Element
Reinforced
Unreinforced
Not permitted
Masonry is effective in
resisting moment and
shear. The cover over the
reinforcement on both
surfaces of the wall
remains intact. System
should essentially remain
elastic.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.003.
Superficial
LOP IV
Significant spalling and
scabbing of masonry.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.003.
Significant spalling and
scabbing of masonry.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.007.
Masonry is completely
disengaged.
Significant spalling and
scabbing of masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry exceeds 0.010.
The concrete cover over the
reinforcement on both
surfaces of the element is
completely disengaged.
Equal tension and
compression
reinforcement that is
properly tied together are
required to resist moment
Limited spalling and
scabbing of masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry should not
exceed 0.010.
Masonry is crushed and not
effective in resisting
moment; however, shear
resistance remains though
interlocking. Compression
reinforcement equal to the
tension reinforcement is
required to resist moment.
The cover over the
reinforcement on both
surfaces of the wall
remains intact.
No spalling or scabbing of
masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry should not
exceed 0.007.
Not permitted
Hazardous
LOP I
Heavy
LOP II
Moderate
LOP III
Table 16.1 Qualitative Damage Expectations for Masonry
430
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.4 Flexural Response Mode of Reinforced Masonry
In this case, Assumption 1 is used to determine the bending capacity of the masonry. The wall should still be designed so that tensile yielding of reinforcement
occurs before the compressive block of masonry starts crushing. This will result
in a relatively weak but ductile wall.
When air-blast pressures are relatively high (in excess of 10 psi), it may be
necessary to design the reinforced masonry wall with a double curtain of reinforcement, using Assumption 2. In this case, it is desirable to place equal
amounts of reinforcement on each side of the wall. Furthermore, it may be necessary to increase the thickness of the masonry wall to accommodate two curtains
of reinforcement. This can, however, result in a stronger and more rigid wall,
which may become susceptible to diagonal and direct shear failure. Thus, there
FAILURE MODES
431
should be an analysis check, and if necessary additional design iterations, to
confirm that the wall is designed in such a way that it reaches the flexural capacity limit before the shear capacity limit, when subjected to load distribution
similar to air-blast load.
16.2.2 Diagonal Tension Shear
Diagonal tension shear failure occurs when a masonry wall response reaches
the diagonal tension resistance limit before the bending capacity is exhausted.
Figure 16.5 depicts typical deflected shape corresponding to diagonal tension
shear response. Diagonal tension failure exhibits very little ductility capacity
and therefore is brittle in nature. Because of this, diagonal tension failure should
Figure 16.5 Diagonal Shear Response Mode of Reinforced Masonry
432
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
be avoided in blast-resistant masonry walls. One method to avoid diagonal tension failure mode is to decrease the flexural capacity of the wall in such a way
that the wall undergoes flexural yielding before shear capacity is exhausted.
However, the flexural resistance of the wall should not be reduced when doing so will compromise the desired level of protection. Alternatively, the diagonal tension capacity of the masonry wall can be increased by increasing the
thickness of the masonry, by grouting more cells, or by placing shear reinforcement. Placement of shear reinforcement is feasible only with two curtains of
reinforcement. If shear reinforcement is required, it should be placed in such a
way that it crosses expected diagonal shear crack between the two curtains of
reinforcement.
Diagonal tension capacity for out-of-plane bending of reinforced masonry
walls can be assessed using ACI-350-08 strength design procedures.
16.2.3 Direct Shear
Direct shear occurs when explosion occurs very close to the wall and when mortar joints are too weak to provide shear stress flow continuity across the masonry
units. Figure 16.6 depicts typical deflected shape consistent with diagonal shear
response. Direct shear may be of concern when scaled distance is less then 4. At
this standoff it is expected that air-blast pressures will be of very high magnitude
and extremely short duration. If the air-blast pressures are of extremely short duration, there will be no time for the masonry wall to respond in flexure mode,
and direct shear response will dominate.
Direct shear capacity depends on friction resistance across the mortar joint
and dowel action of the longitudinal reinforcement. Direction shear strength generally exceeds diagonal shear strength. The direct shear failure mode is almost
always very brittle and should be avoided. One way to avoid direct shear failure
is to increase the standoff between the masonry wall and the expected explosion
source. Furthermore, since reinforced masonry walls do not perform well under
extremely high air-blast pressures, direct shear failure modes should be avoided
whenever possible.
16.2.4 Breach and Spall Phenomena
Breach occurs when a weapon is placed directly against or very close to the
masonry wall surface. Detonation at this close proximity causes shattering of
the masonry units. Breach analysis is often conducted using computational fluid
dynamic codes with appropriate equations of state, or by experimental studies.
Breach effects can be mitigated by masonry wall thickness, proper confinement,
and application of anti-spall laminates such as fiber-reinforced polymer (FRP).
Spall typically occurs on the back side of the wall when a weapon is
placed at the breaching distance but does not cause full breach. Spall occurs
due to a compression wave traveling through the thickness of the masonry.
FAILURE MODES
433
Figure 16.6 Direct Shear Response Mode of Reinforced Masonry
When reflected from the back surface of the wall, this causes through-thickness
tensile stresses at the back face that can exceed the tensile strength of the
masonry.
Both breach and spall will cause fragmentation on the back side of the wall.
Spall can be mitigated by application of shielding (laminate material) such as
FRP or so-called “catchment systems” such as geotextile fabric to stop fragments. Laminate shielding is permanently adhered to the wall surface (on the
protected side of the wall) and may act compositely with the masonry, while
catching systems are basically a curtain or fabric located on the protected side
of the wall to catch debris generated by failed masonry. Both systems provide protection by the development of membrane resistance, and therefore they
should be connected to supports (normally floor slabs) in a way that allows full
434
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
development of the membrane capacity of the base material. This is done to avoid
premature failure of the system. Fragmentation effects are discussed in the detail
in Chapter 8 of this book.
16.3 REINFORCED MASONRY DETAILING
The detailing guidelines presented in this section are intended to promote the
desired structural behaviors discussed in Chapter 9.
Since masonry walls resist air-blast loads primarily though nonlinear deformations, it is important that wall supports develop the full flexural capacity of
the wall. Reinforced and unreinforced walls can be analyzed using single-degreeof-freedom (SDOF) methods or pressure-impulse (P-I) diagrams. Axially loaded
walls should be designed considering slenderness effects.
Non-load-bearing partition walls should be reinforced and grouted, as required, to resist the applied blast loads. Alternatively, debris catch systems
should be used to minimize the debris impact hazard. The effectiveness of the
debris catch systems should be determined through applicable explosive test data
or advanced finite element analysis.
The design shear forces should not be less than the shear forces associated
with the ultimate flexural strength of the element. The shear capacity of reinforced masonry wall structures can be increased by placement of shear reinforcement. Shear reinforcement should be detailed in accordance with UFC 3-340-02.
Masonry wall structures should be sufficiently reinforced to resist the in-plane
and out-of-plane response to dynamic blast loads. It should be permissible to
consider arching action for the evaluation of wall response and to use empirical P-I charts based on masonry wall tests; however, the impulse asymptote
“layover” (the “bump” at the extreme asymptote) associated with the beneficial
effects of negative phase loading may be nonconservative, and should not be considered for design. The use of P-I charts should be limited to flexural modes in
response to external blast loads and to wall types that are within the P-I database.
Diagonal tension and direct shear capacity must be verified independently. Walls
should be designed to develop the shear forces associated with plastic hinging.
Connections should be designed, detailed, and constructed such that the governing failure mode of the connection is a ductile mode related to yielding of the
steel elements of the connection, such as angles, plates, and rebar. Fracture of
bolts and welds should not be the governing failure mode of connections.
This section applies to non-load-bearing masonry walls only. Reinforced concrete masonry units should allow full development of the reinforcing steel. Reinforced concrete masonry unit walls should be designed as either one- or two-way
systems using cell and joint reinforcement. Shear resistance can be provided by
the concrete masonry unit in combination with steel reinforcement, although the
inclusion of this behavior should be considered in light of the assumptions identified at the beginning of this chapter.
REINFORCED MASONRY DETAILING
435
16.3.1 General
All concrete masonry units should be fully grouted for LOP III and LOP IV.
All lap splices should be tension lap splices, as specified in ACI 530 and ACI
530.1 (American Concrete Institute 2008).
Use of mechanical splices should be limited to zones that remain elastic during blast response, unless the mechanical splices can be shown to develop the
expected tensile strength of the bar under conditions generated by blast loads.
All mechanical splices should meet or exceed the specifications in ACI 530 and
ACI 530.1.
Use of welded splices should be limited to zones that remain elastic during
blast response, unless the welded splices can be shown to develop the expected
tensile strength of the bar under conditions generated by blast loads. All welded
splices should meet or exceed the specifications in ACI 530 and ACI 530.1.
16.3.2 Longitudinal Reinforcement
For air-blast-resistant design, it is desirable to place longitudinal reinforcement in each cell in fully grouted concrete masonry units. This will result in
reinforcement placed at 8 in. o.c. This detailing approach is beneficial since every grouted cell adds weight, rigidity, and shear strength, and closely spaced bars
may reduce the size of fragments should the wall fail under air-blast loads. Figures 16.7 though 16.10 provide sample detailing options for reinforced masonry.
16.3.3 Horizontal Reinforcement
A bond beam course with two horizontal reinforcing bars should be placed all
around the cap course of the wall. Bars should be spliced and the course grouted,
per requirements of ACI 530.
All new construction should have a bond beam course with a minimum of two
horizontal reinforcing bars. Reinforcing bars should be located at the level of the
lowest attachment of the masonry wall to the floor diaphragm. Bond beam should
be placed at any horizontal diaphragm or floor system supported by the wall. If
the attachment band in the diaphragm is thicker than the bond beam course, then
the bond beam course should be placed on the bottom of the diaphragm, and
all cells in courses to the top of the attachment band should be fully grouted.
Bars should be spliced and the course grouted per requirements of ACI 530. See
Figure 16.11 for examples of bond beam reinforcement.
A bond beam course with two horizontal reinforcing bars should be placed
at all lintel locations; the bar should be either hooked around vertical reinforcement on either side of the opening or continuous along the wall, with the course
grouted per requirements of ACI 530.
Control joints for all LOPs should be designed, at a minimum, with one
vertical bar in a fully grouted cell column on each side of the control joint.
436
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.7 Fully Grouted Reinforced Masonry Wall
Figure 16.8 Singly Reinforced Masonry Wall
Figure 16.9 Masonry Wall Doubly Reinforced by Staggering Longitudinal Reinforcement
REINFORCED MASONRY DETAILING
Figure 16.10 Doubly Reinforced Masonry Wall – Two Bars per Each Cell
(a)
(b)
Figure 16.11 Examples of Bond Beam
437
438
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Code-approved control joint block and expansion material systems should be
used as necessary in all blast-resistant masonry systems.
The design professional should determine the appropriate locations for control
joints, to the extent that their locations impact blast resistance.
The design professional should specify continuity and/or end conditions of
horizontal bars and bond beams at control joints.
16.3.4 Walls
All elements constructed to LOPs I and II should have at least one continuous
vertical reinforcing bar in a fully grouted cell column, with spacing not to exceed
48 in. o.c.
All elements constructed to LOP III and IV should have at least one continuous vertical reinforcing bar per block in a fully grouted cell column, with spacing
not to exceed 16 in. oc.
In addition, at least one vertical bar should be placed in a fully grouted cell
column at all inside or outside corners, alongside any openings, and in the cell
column on each side of control joints, if they are used in the structure
All elements constructed to LOP III and IV should have a minimum vertical
reinforcement ratio of 0.0025.
Bearing Walls Bearing walls are subjected to constant axial load in addition to
blast loading. The presence of axial load in reinforced masonry increases bending
capacity up to a so-called balanced point on a P/M diagram, after which axial
load causes rapid deterioration of bearing wall air-blast-resistance capacity, due
to buckling. Additionally, the presence of axial load induces larger stresses on
block masonry under compression, and may cause the masonry wall to fail in
concrete crushing mode rather than through the yielding of the reinforcement
(i.e., compression-controlled section).
Careful attention must be paid during the structural analysis of bearing walls
to avoid these undesirable failure modes (crushing and buckling).
Slender Walls Slender walls normally carry some axial load, which affects the
overall resistance of the wall and should be accounted for in the analysis. All
slender wall design should follow ACI 350-08, Section 3.3.5
Infill Walls Infill masonry walls typically act as partitions and are supported by
floor slabs or beams. If properly designed they can provide physical separation
and protection against air-blast waves and fragments.
16.3.5 Support Connections
Since masonry walls are primary blast-load-resisting systems, and since masonry
walls are normally expected to respond to air-blast load in nonlinear regime,
UNREINFORCED MASONRY
439
it is necessary to provide sufficient support to allow masonry walls to develop
full plastic mechanism. This can be achieved by fully developing longitudinal
reinforcement into the supports.
16.4 UNREINFORCED MASONRY
Unreinforced masonry, due to its brittle nature, is not recommended in new construction. However, there is a vast amount of unreinforced masonry in existing
and historic structures. Sometimes, existing reinforced masonry walls contain
reinforcement that provides negligible improvement over the unreinforced masonry. Lightly reinforced masonry walls of this type are usually analyzed using
the principles of unreinforced masonry discussed in this section.
Methods have been established for assessing the stability of unreinforced masonry walls under blast loading. The phenomenon of arching action is one of the
primary mechanisms by which unreinforced masonry resists complete failure.
However, even slight in-plane vertical movements of “rigid” supports negate the
large benefits predicted by the assumed arching action.
16.4.1 Performance Evaluation
When an external explosion imposes an air-blast load on a heavy external wall
system, the air-blast pressure must overcome the inertial forces of the wall itself before putting the system into motion and causing damage. In the case of
thick, unreinforced masonry walls, these forces are very large, and therefore the
resulting damage is less than one might intuitively imagine.
Informative data for masonry can be obtained from the Concrete Masonry
Unit Database Software (CMUDS) program. The full-scale equivalent dimensions for the data have spans of 8 to 11 feet and thicknesses of 6 to 8 inches.
CMUDS also provides data on reinforced masonry, with reinforcement ratios
ranging from 0.15% to 00%, although most of the data have reinforcing ratios from 0.15% to 0.30%. The data include full-scale shock tube testing and
high explosives testing on full- and quarter-scale walls. Elastic-perfectly plastic analysis should not be used for unreinforced masonry elements, as this will
lead to a nonconservative calculation of the blast load that such elements can
resist.
Wall Analysis A common configuration for windows in historic unreinforced
masonry buildings, especially those constructed in the late 1800s and the early
1900s, is to have two or three windows per structural bay. This creates masonry
piers between individual window openings. These piers tend to be critical wallsupport elements.
It is best to begin calculations assuming that the windows can be anchored
to the walls at all four sides, distributing the forces from the windows to the
440
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
supporting walls at the head, sill, and jambs. Walls should have sufficient dynamic capacity to allow window glazing to break first. This approach, commonly
known as balanced design, has been shown to reduce the weight (the size and the
amount) of debris that may enter the occupied space. There is a direct correlation
between injuries and size and velocity of debris.
Two computational models for analyzing wall response are described below.
Simply-Supported Wall A conservative approach to modeling the resistance of
an unreinforced masonry wall for an air-blast load is to represent the wall as a
slender member subjected to a uniformly applied dynamic load. It is assumed
that there is no contribution of axial compressive forces at the top and bottom
of the wall, and the wall is analyzed as a simply-supported beam. The flexural resistance of the wall is dependent on the compressive stresses from its self
weight, and the tensile strength as defined by the mortar’s modulus of rupture.
Because unreinforced masonry has little tensile resistance, the flexural capacity
of the wall is minimal, resulting in extensive cracking and brittle failure at the
interior face of the wall.
Rigid Supports Another model, based on UFC 3-340-02 (the recognized
air-blast-resistance design manual throughout the United States), assumes
that the top and bottom supports are completely rigid and provide restraint
against elongation. Strength comes from compressive blocks that form at the
top, bottom, and mid-height hinge locations of the wall, behavior known as a
compressive membrane phenomenon. This model also assumes that the mortar
joints will be cracked, disregarding what little flexural tensile strength remains
in the masonry. The deflected wall “wedges” itself between floor slabs and also
forms a plastic hinge at mid-height. This model takes into account that masonry
infill walls will often have a gap at the top, either from a mortar joint or from a
design that accounts for flexure of the floor systems above.
16.4.2 Retrofit Recommendations
Retrofit masonry walls can be costly, but may be unavoidable due to various
constraints such as the historic value of a building. Masonry wall retrofits are
usually required when an existing wall cannot be replaced and does not meet the
desired performance level of protection. Retrofit recommendations provided in
this section apply equally to unreinforced and reinforced masonry walls.
Catchment Systems Or Curtains Depending on the performance intent, an alternative to structural retrofit of unreinforced masonry walls is to design geotextile fabric blast curtain located on the inside face of a non-blast-mitigating masonry wall. This curtain will provide protection from masonry fragments. The
geotextile fabric and connections should be designed for membrane action of the
fabric. The fabric connections should transfer loads into the floor system without
UNREINFORCED MASONRY
441
causing fabric tears from stress concentrations. The performance of a specified
geotextile fabric and its connections should be verified by blast testing. Generally, blast testing, in addition to demonstrating acceptable performance for design
loads, should also demonstrate the post-failure behavior of the retrofit system, to
ensure that the failure of the system does not create brittle and hazardous failure.
Secondary Walls A secondary blast-mitigating wall system installed on the
inside face of an existing wall system, in addition to being designed to withstand
fragmentation and impact effects (due to the frangible wall in front), should also
be designed as a new wall, in accordance with this chapter.
Steel Supporting Frames The most straightforward retrofit is to install steel
frames, interior to the existing walls, that support the wall and the window system. This solution is appropriate where the wall piers are not able to support the
additional air-blast load imposed by the windows. This retrofit is often unacceptable to architects, as it either creates an uneven wall surface or requires reducing
the overall room size by furring out the walls.
Shotcrete Walls In some cases, other loading condition requirements may be
combined with air-blast requirements to develop multipurpose solutions. Where
progressive collapse or seismic design requirements would benefit from shotcrete
walls, this may be used as an air-blast retrofit of unreinforced masonry walls as
well. Window systems could be mounted in the new concrete wall system. This
would be an especially useful technique for relatively larger blast loads.
Fiber-Reinforced Polymer (FRP) Another retrofit method is to apply vertical FRP composite strips along the inside face of the masonry wall. FRP
composites have significant tensile strength in the direction of the fibers and
can be highly effective in increasing the wall’s out-of-plane flexural capacity. Additional advantages of this method are that it is relatively unobtrusive
to the architectural detailing of the wall and existing structure, and its application process is not labor-intensive, requiring little disruption to building
occupants.
The flexural capacity of an FRP-retrofitted unreinforced masonry wall can
be assessed using basic moment-equilibrium relations, as with a steel-reinforced
masonry wall. The tensile strength of the masonry is disregarded, and all tensile
resistance is assumed to be provided by the FRP. The flexural resistance can
then be computed as a function of the compressive strength of the masonry, the
tensile strength of the FRP, the thickness of the masonry panel, and the value of
the applied axial force.
Blast tests have demonstrated that it is unlikely that the wall’s predominant
failure mode will be failure of the FRP. Also, tests of unreinforced masonry
units to out-of-plane air-blast loading have demonstrated that shear failure at the
supports is a predominant failure mechanism. The FRP reinforcing provides no
442
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
additional shear resistance to the wall section, and if connections at the wall supports do not allow transfer of excessive shear forces to other structural elements,
shear failure at the wall supports can occur. Also, connections at wall supports
are instrumental in preventing out-of-plane flexural collapse towards the front
side of the wall due to rebound forces.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Masonry Structures
and Commentary (ACI 530-08/ACI 530.1-08/ASCE 5-08/TMS 602-08). Farmington
Hills, MI: American Concrete Institute.
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
ASTM. 2000. Standard Specification for Mortar for Unit Masonry (C270). West Conshohocken, PA: ASTM International.
. 2006. Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement (A706). West Conshohocken, PA: ASTM International.
. 2008a. Specification for Deformed and Plain Billet-Steel Bars for Concrete Reinforcement (A615). West Conshohocken, PA: ASTM International.
. 2008b. Standard Specification for Grout for Masonry (C476). West Conshohocken, PA: ASTM International.
Brandow G. E., G. C. Hart, and A. Virdee. 1997. Design of Reinforced Masonry Structures. Citrus Heights, CA: Concrete Masonry Association of California and Nevada.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5−1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2003. Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks (FEMA 427). Washington, DC: Federal Emergency
Management Agency, Department of Homeland Security.
Paulay, T. and M. J. N. Priestley. 1992. Seismic Design of Reinforced Concrete and Masonry Buildings. Hoboken, NJ: John Wiley and Sons.
Protective Design Center. 2008. Single Degree of Freedom Structural Response Limits for
Antiterrorism Design (PDC TR-06-08). Omaha, NE: U.S. Army Corps of Engineers,
Protective Design Center.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (UFC 3-340-0). Washington, DC: U.S. Department
of Defense, Unified Facilities Criteria Program. For Official Use Only.
Unified Facilities Criteria Program. 2006. DoD Minimum Antiterrorism Standards for
Buildings (UFC 4-010-01). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
Unified Facilities Criteria Program. 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program. Supersedes TM 5-1300.
UNREINFORCED MASONRY
443
U.S. Army Corps of Engineers. 1999. Airblast Protection Retrofit for Unreinforced
Concrete Masonry Walls (ETL 1110-3-494). Washington, DC: U.S. Army Corps of
Engineers.
U.S. Department of State, Overseas Buildings Operations. 2002. A&E Design Guidelines
for U.S. Diplomatic Mission Buildings. Washington, DC: U.S. Department of State,
Overseas Building Operations.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
17
Retrofit of Structural
Components and Systems
John E. Crawford and L. Javier Malvar
17.1 INTRODUCTION
Structural components such as columns, beams, floors, and walls are likely to be
vulnerable to blast loads. This is important because the resulting damage to these
components could produce partial or complete collapse of the structural system,
or produce large volumes of injurious debris. For example:
r The blast at the Alfred P. Murrah Federal Building in Oklahoma City led
to the shearing of main load-bearing columns and resulted in the collapse
of nearly half of the building’s structure, which resulted in a large number
of casualties. In this incident, it was estimated that 87% of the people in
the collapsed portion died (153 out of 175), whereas only 5% of the people
in the uncollapsed portion died (10 out of 186) (Federal Emergency Management Agency 1996), indicating that, had the structure better resisted the
blast loads and not experienced a collapse, the casualties would have been
a lot fewer.
r Even when the structural system survives, blast loads can cause the disintegration of walls and windows, creating secondary fragments that will
generate fatalities and injuries in the adjacent rooms.
In this chapter, methods are described for retrofitting structural components
to enhance their blast resistance and for assessing the protection the retrofits
provide.
Much of this chapter is devoted to describing retrofit techniques to protect
reinforced concrete (RC) columns. This was done because so much can be done
to prevent the catastrophic damage observed for some types of RC columns when
subjected to blast. This information, along with that for the other components, is
presented in a summarized form due to space limitation.
445
446
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
17.2 RETROFIT OF COLUMNS
Typically, columns are the component whose integrity is key to sustaining the
capability of a building’s structural system to survive a blast load. As such, attention to ensuring their survival should have the highest priority when the blast
resistance of a building is to be upgraded.
17.2.1 Reinforced Concrete Columns
Soon after the bombing of the Alfred P. Murrah Building, it was shown that highfidelity physics-based (HFPB) finite element (FE) models would have predicted
the failure of the bearing columns nearest the point of detonation, and the resulting progressive building collapse (Crawford, Wesevich, et al. 1995; Crawford,
Malvar, et al. 1996; Crawford, Bogosian, and Wesevich 1997). These models
also suggested that the use of composite wraps, or steel jackets, could have increased the columns’ blast resistance sufficiently to prevent their failure, and
thus, the building’s collapse.
The concept of jacketing columns had also been suggested by the Federal Emergency Management Agency (Federal Emergency Management Agency
1996), which indicated that retrofit techniques used to resist seismic loads could
be extended to resist blast loads. But it was not until 1999 (Crawford 2001) that,
in a series of three blast tests conducted by the Defense Threat Reduction Agency
(DTRA), it was demonstrated that steel and fiber-reinforced polymer (FRP) jacketing could be used to significantly enhance the blast resistance of RC columns.
Test Data These first three blast tests were conducted on columns located on
the ground floor of a full-scale four-story RC building (Figure 17.1); they are
designated as Tests 1 to 3 in the figure. The same blast load on identical columns
(i.e., except for the retrofit) was used in all three tests, the results being shown in
Figure 17.2.
The Test 1 column’s performance under these conditions is shown in Figure
17.2a, where a diagonal shear failure can be observed—which is not surprising, given the relatively wide tie spacing of this conventionally designed column. However, when this type of column design is retrofitted with a steel jacket
(Test 2), or carbon FRP (CFRP) hoop wraps and vertical CFRP strips (Test 3),
it can survive the same load in nearly pristine condition (Malvar, Morrill, and
Crawford 1999).
Based on these initial results, an extensive series of full-scale column components was tested using various types of FRP retrofit designs. The results from the
blast tests were augmented with a companion program of ten laboratory tests of
column specimens identical to those tested by blast, but loaded statically with hydraulic rams (Figure 17.3). These rams were applied in such a way as to approximate the deformation patterns observed in the blast tests. Aramid FRP (AFRP),
glass FRP (GFRP), and CFRP composites were used successfully in these tests,
although CFRP has shown better durability characteristics (American Concrete
RETROFIT OF COLUMNS
447
Figure 17.1 Location of initial blast tests of unretrofitted and retrofitted RC columns
(tests were conducted using a four-story building constructed using a cast-in-place conventionally designed RC flat-slab framing system; figure shows structure after Test 1
(Photo courtesy of DTRA)
Institute 2002, Malvar 1998a), and its higher stiffness can provide more efficient
confinement with less material (Malvar, Morrill, and Crawford 2004).
Design/Analysis Tools One of the major outcomes of the test program was
the development and validation of a procedure for the design of FRP wrap and
steel jackets to retrofit RC columns to enhance their blast resistance. This procedure can quickly assess the performance of both conventional and retrofitted
RC columns subjected to blast loads, and it provides an easily used means to
design an FRP wrap or steel jacket to retrofit an RC column so that it can survive
a specific blast threat (Malvar, Morrill, and Crawford 1999; Morrill, Malvar, and
Crawford 1999; Morrill, Malvar, and Crawford et al. 2000; Morrill, Malvar, and
Crawford et al. 2004; Perea and Morrill 2002; Kersul and Sunshine 2002). The
retrofit design procedure was embodied in the code CBARD (Crawford, Malvar,
and Ferritto et al. 2003). For composites, the design procedure determines the
number of FRP hoop wraps needed (and possible vertical FRP reinforcement)
for a given blast load. For steel jackets, it determines the required jacket thickness, given the proposed jacket geometry.
448
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
(a)
(b)
(c)
Figure 17.2 Depiction of results from full-scale blast testing of retrofitted and unretrofitted RC columns (Photo courtesy of DTRA): (a) Unretrofitted column (Test 1),
(b) Steel jacket column (Test 2), (c) FRP wrapped column (Test 3).
RETROFIT OF COLUMNS
449
Figure 17.3 Quasi-static laboratory test demonstrating the highly ductile behavior
achievable for an RC column retrofitted with CFRP wrap to prevent shear failure (Photo
courtesy of DTRA)
450
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.4 Measured versus predicted resistance function for an RC column (as computed by CBARD), which is given in the form of lateral pressure applied versus lateral
deflection at mid-height (the added capacity beyond the third hinge formation is provided
by compression-membrane behavior)
In addition to the blast tests, as mentioned, static laboratory tests (Figure 17.3)
with loading conditions that simulated the deformations observed in the blast
tests were conducted to measure the flexural resistance of the columns (Morrill
et al. 2001). Static testing cannot replicate the part of the behavior related to
the initial highly transient response driven by the short-duration pressure loading of the blast (e.g., the apparent material strengthening at high strain rates),
or the wrapping of the blast loads around the column, but it can yield data that
can be used to verify the blast design procedure. For example, these data provide demonstrative evidence of the importance of compression-membrane (C-M)
forces (Figures 17.4 and 17.5), demonstrate the effectiveness of FRP wrap at enhancing shear resistance (Figure 17.5), and provide load-deflection information
for validating design tools (e.g., CBARD).
Although retrofitting columns with steel jackets or composite wraps is reminiscent of seismic retrofits, there are basic differences. For example, apparent
material strengthening occurs at the high strain rates induced by the blast loading (Malvar 1998b, Malvar and Ross 1998), a phenomenon that, for concrete,
has been linked to moisture content at lower strain rates (below about 1/s) and to
inertial effects above that, and that is typically neglected during seismic loading.
Also, the deformed shape of building columns under blast loads usually includes
(a)
(b)
Figure 17.5 Response of conventional and CFRP hoop retrofitted 14-inch square RC
columns (the axial load in the CFRP column initially increases with increasing lateral
deflection due to compression membrane then decreases to 100 kips when the test was
stopped): (a) Axial load, (b) Lateral load
452
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.6 Classic form for resistance function (represented by the middle line) pertaining to the fully ductile response of an RC column; this resistance only develops if the
diagonal shear capacity always exceeds it
two inflection points and three plastic hinges (Figure 17.3), as compared to a
single inflection and two plastic hinges for seismic loading. This is an important difference, as the deformed shape with two inflexion points allows for the
development of C-M forces, which can significantly increase the lateral resistance of the column, typically anywhere between 20% and 100%, depending on
the column’s geometry and the rigidity of the boundary conditions (Crawford,
Malvar, and Morrill et al. 2001). This C-M behavior develops due to the vertical
(axial) constraint at the top of the column provided by the inertia of the supported structure under high loading rates. Figure 17.6 shows schematically the
flexural resistance after the formation of three hinges and the increase provided
by the C-M resistance developed subsequently. C-M behavior is typically not
present during seismic events; it is a phenomenon associated with both the form
and rapidity of the blast response and the presence of significant mass above the
column.
Procedure for Retrofitting RC Columns The key concept employed in
retrofitting an RC column pertains to preventing shear and other brittle forms of
failure under the blast loading, by ensuring that the shear capacity of the column
exceeds its flexural strength (i.e., including its C−M capacity), and therefore
RETROFIT OF COLUMNS
453
ensuring that the column dissipates the most energy possible via flexure. For example, in Figure 17.6, if the shear capacity is depicted by the top line (demarked
with the solid squares), then the complete flexural behavior will be allowed to
develop. This latter situation ensures that the brittle shear failure response is
avoided altogether (i.e., over the whole range of the column’s response), and instead ensures a ductile flexural response up to the point where P-δ failure occurs.
When the shear capacity of the column is too low (e.g., if in Figure 17.6 the shear
capacity is represented by the lowest line, demarked by the open triangles), this
can be remediated by employing steel jackets or composite wraps to shift this line
upward—for example, to the location of the high shear resistance line. In the case
of composites, hoop wraps in general are sufficient to provide the required shear
capacity and ensure ductile behavior, and vertical sheets or strips are not needed.
Moreover, vertical composite reinforcements tend to rupture at low column lateral displacement, often before compression membrane is developed—that is,
at a deflection of half the column depth, or less; see DSWA (Defense Special
Weapons Agency 1998) or Malvar and Wesevich (1996)—whereas the hoop
wraps can remain intact during large lateral deflections (see Figure 17.3).
Description of Design Methodology. This methodology contains a number
of unique features that distinguishes it from other design procedures, such as
those for retrofitting columns to improve their seismic capacity and traditional
American Concrete Institute (ACI) design. These features include:
r Recognition that some RC columns, even those designed for regions of high
seismicity, are prone to fail in shear and/or experience catastrophic loss of
axial capacity under a blast load.
r Using resistance functions (e.g., as shown in Figure 17.4) specially tailored
to the column’s geometry and the retrofit concepts employed.
r Focusing the retrofit design on achieving a fully ductile response mechanism (e.g., as demonstrated in Figure 17.5) for the lateral response of the
column. A primary goal in designing the retrofit is to have a column that is
fully ductile under lateral load.
r Using a best-estimate response-prediction model for selecting design parameters and estimating capacities of retrofitted and unretrofitted columns.
r Using a single-degree-of-freedom (SDOF) model to predict responses to
blast loads for both unretrofitted and retrofitted RC columns.
Blast loads may produce several types of response indicating various deficiencies, or lack thereof, in the column’s behavior. The deficiencies likely to be
encountered include:
r Lack of Diagonal Shear Capacity: The column can fail in shear with little
plastic behavior, hence dissipating little energy. This is a worst-case condition for a column subjected to a blast load, and will likely be the predominant reason for retrofitting columns. This deficiency is sometimes referred
454
r
r
r
r
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
to, although not preferably, as diagonal tension failure rather than diagonal
shear failure.
Lack of Flexural Ductility: Good practice would dictate that columns should
achieve a full measure of flexural ductility for blast-resistant design—a full
measure being reaching their deflection limit, which is related to keeping
the column’s axial demand/capacity ratio less than 1 (one). Deflections associated with this level of ductility (i.e., full measure) are typically on the
order of a large fraction of the column’s width. This may not be practical for
wide, short columns, in which case, one must be content with prevention of
shear and other brittle failure modes.
Lack of Direct Shear Capacity: This condition is uncommon for building
columns. It occurs for columns with high diagonal shear capacity and high
flexural resistance. This condition is of particular concern for columns having low ratios of height to depth, which is common in parking structures
underneath multistory buildings, and for columns with steel jackets. It is
often not practical or easy to enhance the direct shear capacity, but significant energy can be dissipated by the direct shear response itself so that an
acceptable behavior may be achieved without resorting to any extraordinary
retrofit measure.
Lack of Moment Capacity: If the resistance function is low or has little area
under it (i.e., its integral is small, indicating little energy absorption in the
response), little energy will dissipate, and unacceptable lateral deflections
may ensue. This may be mitigated by adding vertical composite strips (FRP
retrofits), a steel jacket, or steel strips (steel retrofits). In these cases, close
attention should be paid to ensuring that the resulting enhanced diagonal
and direct shear demands are met.
Lack of Confinement: This condition is of most concern for large columns
having high axial loads that are subjected to a large blast load. In such a
situation, the lateral response to the blast is likely to be quite small, but the
shock load may be large enough to cause extensive cracking in the concrete.
In this case, if the column has insufficient ties to maintain confinement of
the concrete and prevent rebars buckling, then the axial resistance of the
column may be severely compromised.
The design methodology presented in CBARD encompasses these situations,
and provides a means for designing either FRP wrap or steel jackets to address
all of these deficiencies.
17.2.2 Steel Columns
The retrofit of steel columns is relatively simple, as compared to the design process employed for RC columns. Generally, the main concerns for steel columns
are that they are likely to lack sufficient shear capacity near their supports, and
that they may not have enough flexure resistance to preclude excessive lateral
RETROFIT OF COLUMNS
455
deformation. Retrofit concepts for blast can often employ concepts similar to
those for enhancing the static load resistance of a steel column.
There are considerable blast effects data for steel columns and subassemblies comprising a column fixed at its base with a set of girders framing into
it at a floor level above, and the connections used (Crawford, Magallanes, and
Morrill 2005). These data have provided insight into the behaviors that may
be cause for concern in a blast environment and that may require mitigation.
The data have also been used to validate HFPB finite element models and simplified engineering models for the analysis and design of steel framing systems for blast loads (Morrill, Crawford, and Magallanes 2007, Magallanes and
Morrill 2008).
Damage Modes. In retrofitting steel columns to enhance their resistance to
blast loads, three basic modes of response should be considered:
r Mode 1: Excessive lateral deformations
r Mode 2: Joint failures, for example, for baseplate connections and column
splices that are near the point of detonation
r Mode 3: Fracture or excessive localized deformation of web and flange
The first two modes are relatively straightforward in terms of using retrofit
concepts to mitigate the risk of their occurrence, which are primarily related
to the lateral loading engendered by the blast. The Mode 3 failure, however,
is peculiar to blast where the rapid loading produced by the blast produces a
shear failure along the K-line, or causes the web and flanges to deform excessively. Results from a blast test illustrating some aspects of these failure
modes, taken from one of a series of tests of wide flange segments, are shown in
Figure 17.7.
Retrofit designs for addressing these potential failure mechanisms are likely
to require careful consideration, and no presupposed methodology or approach
can be considered adequate to ensure that all the salient issues at a particular site
are addressed adequately.
Retrofitting. Mode 1 failures can be addressed in relatively conventional
ways by strengthening of the member (e.g., by the addition of cover plates).
Both local and global deformation of the section may need to be addressed, depending on the blast scenarios considered. Concerns for excessive deformation
for light sections may be particularly difficult to address. The retrofit approach
would follow much the same strategy as would be used to retrofit a girder to
take a higher load. This might involve the use of cover plates, stiffeners, knee
braces, and gusset plates. The main concern in developing the retrofit design
will be ensuring that ductile behavior is maintained throughout the response,
especially with regard to the added shear demand caused by blast type loadings at the column’s supports. Alternatively, a blast shield may be installed in
front of the column to reduce its response and negate or lessen the need for
retrofit.
(a)
(b)
(c)
Figure 17.7 Examples of failure modes for steel columns: (a) Section view, illustrating
Mode 1 P-δ failure, (b) Connection view, illustrating concerns with fracture (Mode 3
failure), (c) Baseplate view, depicting a Mode 2 failure (Photo courtesy of DTRA)
RETROFIT OF COLUMNS
457
Addressing Mode 2 failures is a little less straightforward, but they can be
approached in several ways. These failures are related to a general lack of capability in the various sorts of connections used in conjunction with deploying
steel columns. Of these, the failure of column base plates and splices is of most
concern. The retrofits for addressing Mode 2 failures are likely to take a variety
of forms. The general approach employed in reducing the vulnerability to Mode
2 failures is to ensure that the connection does not represent a weak spot. For
example, using a cross tie to remove reliance on the base plate’s capacity. The
other main concern is to ensure that plastic hinges are able to form in the column
as a whole, so as to achieve a fully ductile lateral response for the column.
Mode 3 failures are most easily addressed by taking advantage of the short
duration of the blast load and the benefits that can be derived from inertia in this
situation. A retrofit concept embodying this approach is depicted in Figure 17.8.
Here the interstices between the section’s flanges are filled with concrete, which
mitigates problems associated with localized bending of both the flanges and
webs (Crawford and Lan 2005). This may be augmented with additional webs,
as shown in the figure, if additional shear capacity is desired.
Another means to protect the column from web breach (i.e., in the case of a
wide flange section) is to place a concrete or other similarly substantive panel
in front of the web so that the airblast cannot impinge directly on the web. This
could be in the form of a conventional cladding (Crawford, Magallanes, and
Morrill 2005).
Figure 17.8 Example of the infill concept for preventing web and flange failure
(Mode 3) for steel sections
458
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
17.3 RETROFIT OF WALLS
Bearing walls, like columns, are a key structural component supporting the gravity loads of a building system, and the loss of bearing walls can also lead to overall collapse. Unlike columns, however, bearing walls can sustain major damage
without compromising their function of carrying the gravity loads. These walls,
in essence, offer a continuous form of load path redundancy, which allows them
to be locally breached by a nearby charge and still perform their primary function of holding the structure up. In contrast to columns, walls can also provide
a last line of defense for the building’s occupants from the effects of blast, and
the failure of both bearing and nonbearing walls can result in deadly secondary
fragmentation likely to produce significant casualties as their debris are blown
through a building.
Walls are found in a wide variety of strengths, combine a wide range of materials, vary substantially in quality of construction, and often present considerable
difficulty in estimating their robustness as a structural system. In the following
discussion, the behaviors of walls subjected to out-of-plane static and blast loads
are examined. These are different from seismic applications, where walls are
generally used more to provide shear resistance, and therefore subjected primarily to in-plane loads.
Composites, polymers, and conventional materials have been used to enhance
the blast resistance of masonry walls (Malvar, Crawford, and Morrill 2007).
FRPs have been used to enhance the resistance of RC and stud walls. Sheetmetal panels and blast shields are useful in enhancing the blast resistance of any
wall. Examples of the use of some of these retrofit concepts are described below.
17.3.1 Masonry Walls
The retrofitting of masonry walls with and without reinforcement to enhance
their blast resistance has been widely studied. Those composed of concrete masonry units (CMU) and fired clay bricks will be the focus of the following discussion.
The response of conventional masonry walls to blast loads has been observed
in many full and subscale tests. An example of one of these tests is shown in Figure 17.9. In this figure, two side-by-side identical ungrouted CMU walls were
tested with a moderate-size blast load. The wall to the right was retrofitted (Figure 17.9b); the one to the left was not. Figure 17.10 shows a snapshot depicting
the motions of the unretrofitted wall immediately after a detonation. The performance shown is typical of a masonry wall’s response for a moderate blast load,
in that the CMU blocks remain relatively intact, while the failure of the wall is
largely due to the opening of the masonry’s horizontal joints.
Three approaches to retrofitting walls are described. The first two technologies involve bonding reinforced and/or unreinforced layers of polymer to the
back (and possibly front) faces of a masonry wall to provide both added strength
and ductility for blast load response. The third approach uses some of the same
(a)
(b)
Figure 17.9 Example showing the benefit of using FRP to retrofit a masonry wall to
enhance its blast resistance: (a) Control (left) and retrofitted wall (right) prior to blast
test, (b) Control (left) and retrofitted wall (right) after the test (Photo courtesy of The
Energetic Materials and Research Testing Center)
460
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.10 Example typical of the response expected from a CMU wall struck by a
blast; at the time shown, the CMU is moving at several tens of feet per second (Photo
courtesy of The Energetic Materials and Research Testing Center)
materials as the first two, but in this case they are not bonded to the wall. In this
situation, the strengthening elements, such as FRP, are anchored to the floor and
act by preventing entry of wall debris rather than by wall strengthening.
Fiber-Reinforced Polymer The use of FRP to enhance the strength and blast
resistance of masonry walls has been extensively studied. This would include
FRPs such as CFRP, AFRP, and GFRP.
CMU Studies. Muszynski (1998) and Muszynski and Purcell (2003) bonded
CFRP and AFRP to masonry walls composed of concrete masonry units. Baylot, Dennis, and Woodson (2000), Baylot et al. (2005), and Dennis, Baylot, and
Woodson (2002) conducted experiments on walls using quarter-scale models
built with typical 8-inch-wide CMU. In these tests, three types of retrofitting
schemes were evaluated for ungrouted and for partially grouted walls. The first
retrofit method consisted of 0.04-inch (1 mm) thick GFRP fabric bonded to
the back face of the wall. The second retrofit technique consisted of a twopart sprayed-on polyurea product applied to the back face of the wall. The third
RETROFIT OF WALLS
461
retrofit method used 20-gauge galvanized steel sheet metal attached to the back
of the wall. All the walls failed during the blast testing by separating from the
frame, but the GFRP and the polyurea retrofitted walls were successful in preventing loose fragments within the structure, and were, therefore, considered as
successful retrofits.
Myers, Belarbi, and El-Domiaty (2003, 2004) tested eight ungrouted, unreinforced masonry (URM) walls retrofitted with FRP (epoxy) laminates and
near-surface-mounted (NSM) rods. Two series of walls (l00 and 200 mm or 4
and 8 inches thick) with different slenderness ratios and different strengthening
schemes were tested at varying blast charge weights and standoff distances. The
walls were only supported at the top and bottom, and retrofitted walls failed in
shear at those locations. The FRP laminates were able to contain the wall debris.
Carney and Myers (2003, 2005) tested twelve ungrouted URM walls using a
pressure bag to apply a uniform loading (Phase I), then field-tested two walls under blast (Phase II). Results of the last four Phase I tests, together with the Phase
II tests, are shown in the second paper (Carney and Myers 2005). The walls
were reinforced with GFRP sheets or NSM rods, with and without anchorage.
The anchorages consisted of wrapping the sheets around NSM rods embedded
into the wall’s top and bottom support for the first case, and of extending the
NSM rod three inches into the wall support for the second. Strengthening the anchorages almost doubled the capacity of the walls. In contrast, unanchored NSM
rods did not show any improvement. For FRP reinforced walls, no shear failure
was observed during the static tests, while it was apparent in previous blast tests
(Myers, Belarbi, and El-Domiaty 2003), indicating a potential limitation in the
use of static tests to simulate blast tests.
Stanley, Metzger, and Martinez et al. (2005) used a retrofit system composed
of multiple layers of carbon and glass FRP in an epoxy resin (applied with paint
rollers), placed on the interior and exterior surfaces of the CMU wall test specimen. This system was successful in containing debris but seemed too labor intensive to apply. Stanley, Metzger, and Morrill et al. (2005a) used two layers of a
glass FRP in an epoxy resin applied to the interior surface of the CMU wall test
specimen. This wall experienced complete failure, indicating that stiff retrofit
systems fail catastrophically once their capacity is exceeded.
Simmons (2005) studied the effectiveness of three types of retrofits (E-glass,
22-gauge steel, and an elastomeric polymer) on CMU walls to increase their
resistance to primary fragments from a 120-mm mortar round and a 122-mm
rocket; he found that the E-glass significantly reduced the number of fragments
that passed through the wall.
Brick Studies. Conley and Mlakar (1999) used CFRP (epoxy) strips to reinforce small brick walls loaded with a gas-loaded piston and subjected to one-way
bending. Increases in load capacity of 400% were recorded.
Patoary and Tan (2003) completed static tests using air bags on four built-in
solid masonry brick walls retrofitted with three types of FRP systems: glass, carbon, and low-strength fiberglass woven roving. The walls, 120-mm thick, were
able to sustain deflections up to 20 mm. Specimens with GFRP or CFRP failed
462
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
by edge debonding, while specimens with roving failed by fiber rupture. This
work was followed by out-of-plane tests on 30 masonry walls reinforced with
three different GFRP and CFRP systems (Tan and Patoary 2004).
Retrofitting with Unreinforced or Fiber-Reinforced Polyurea-like Polymers
In some situations, the performance of walls reinforced with FRP composites
is limited by their peak strain capacity, so materials with higher failure strains
might be expected to perform better. Because both fibers and epoxies used for
FRP applications have rather low strains to failure, several researchers started
looking at polyureas, which are materials with purported failure strains of 100%
or more.
These materials have appeared in two forms: one as a spray-on coating that
is applied and bonded directly to the wall’s surface, the other as a sheet polymer
that is placed behind the wall and anchored to the floors. The first system acts in
concert with the masonry by adding tensile capacity. The second system acts as
a catcher system to contain the wall’s debris and is not bonded to the wall. This
latter system is described in Section 17.3.1.3.
Unreinforced. Knox et al. (2000) sprayed the back of concrete block walls
with various unreinforced polymers, in particular polyurethane and polyurea.
These polymers showed exceptional ductility and were able to prevent injurious
wall fragmentation. As indicated earlier, Baylot, Dennis, and Woodson (2000)
and Baylot et al. (2005) and Dennis, Baylot, and Woodson (2002) conducted
experiments on CMU walls using quarter-scale models of typical 8−inch-wide
CMU. Among other retrofits, they used a two-part sprayed-on polyurea product
applied to the back face of the wall, which was successful in preventing loose
fragments within the structure.
Hammons, Knox, and Porter (2002) expanded their application of polyurea
to CMU walls and found it very successful in preventing deadly wall fragmentation. Deflections of up to 610 mm (24 inches) were recorded (the nominal CMU wall width was 8 inches, or 200 mm). Other tests, using a high
pressure spray–applied two-component polyurea (Stanley Martinez, and Koenig
2005b), and a two-component spray-applied polyurea liner (Stanley, Metzger,
and Morrill et al. 2005b) applied to both interior and exterior CMU wall faces,
confirmed such findings, although some tearing of the polymer was observed. In
these applications with large deflections, the complete resistance function may be
mobilized, including the lower resistance beyond compression membrane (due
to the crushing of the concrete compression zone), and the subsequent strengthening due to tensile membrane, enabling the dissipation of significant amounts of
energy.
Davidson, Fisher et al. (2005) studied 12 polymer-reinforced masonry walls
during seven explosive tests. The walls were restrained at the top and bottom
only and responded in a one-way fashion. During the tests, arching due to end
constraints was suspected to have contributed to the loss of the compression face
near the ends. A simple spray overlap of 6 inches (152 mm) was sufficient to
transfer the loads to the frame. Compared to stiff FRP fabric reinforcing, the
RETROFIT OF WALLS
463
spray materials were cheaper, bonded better to the wall, and were easier to anchor
to the frame (see also Davidson, Porter et al. 2004).
Reinforced. These polyurea polymers can also be further strengthened with
continuous fibers, which can provide some additional energy dissipation up to
fiber rupture. Tests were conducted by Crawford and Morrill (2002) to define
the properties of reinforced and nonreinforced polyurea spray-on coatings.
These materials were used to retrofit masonry walls subjected to blast loads.
This is detailed in several studies (Crawford and Morrill 2003a, 2003b; Johnson,
Slawson, Cummins et al. 2003, 2004; Johnson and Slawson 2004; Morrill,
Nicolaisen, and Hutchinson 2005; Piepenburg et al. 2002, 2003; Stanley,
Metzger, and Morrill et al. 2005c; Stanley, Metzger, and Koenig. 2005a).
Piepenburg et al. (2002) used an aramid grid fabric to reinforce a polyurea
spray retrofit of a CMU wall, and were successful in containing the CMU debris. Piepenburg et al. (2003) also used an aramid within two thin sprayed-on
polyurea layers on a CMU wall. Crawford and Morrill (2003a) describe a retrofit
consisting of an aramid laminate that is secured to the wall with a polyurethane
adhesive, and anchored to the diaphragms with energy-dissipating compliant anchorages. Johnson, Slawson, and Cummins et al. (2003, 2004) used an openweave aramid with different fiber lay-ups with a spray-on polyurea. The capacity
of the walls retrofitted with reinforced polyurea was 1.4–2.0 times higher than
that of the walls retrofitted with just polyurea, and 5.5–7.5 higher than that of the
unretrofitted walls. Morrill, Nicolaisen, and Hutchinson (2005) used a polyurea
coating, with and without an aramid fiber mesh, on the back side of brick wall
samples tested quasi-statically in a laboratory environment. These tests were useful in determining complete quasi-static resistance functions, including tensile
membrane.
Stanley, Metzger, and Morrill et al. (2005c) used a spray-applied two-part
polyurea with an inner aramid weaved fabric on a CMU wall, and successfully
stopped the CMU debris. Peak deflections were about 230 mm (9 inches). Stanley, Martinez, and Koenig (2005a) used a non-woven polypropylene geotextile
fabric in between two layers of sprayed-on polyurethane applied to the interior
and exterior surfaces of the CMU wall test specimen. In this case, the retrofitted
wall did fail, although it would appear that the retrofit had provided some amount
of early-time confinement and lateral restraint of the CMU blocks.
Retrofitting with Catcher Systems In contrast to the retrofit schemes described
for walls above, which are related to strengthening them directly, catcher
systems are placed behind the wall to catch any debris that might be generated
by its breakup from a blast. Such a device may be composed of thin sheets
R
of fabric (e.g., Kevlar laminate), sheet metal, or polymer (e.g., as shown in
Figure 17.11).
This form of catcher system may be augmented with cables for cases involving particularly high loads (e.g., impulses of thousands of psi-ms) or where
anchorage conditions prohibit the use of the sheets by themselves. These systems
are designated as cable catcher systems.
464
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.11 Example of a thin polymer catcher system being installed (Photo courtesy
of Sherwin-Williams Company)
The intent of these concepts is just to capture the debris of the wall, not to
strengthen the wall. These systems may be constructed in a variety of ways, and
some designs have very large capacities. In addition, these systems are likely
to impart less lateral load to the structural system of the building, which is
significant where collapse of the framing system due to the lateral loading of
the blast is of concern. The debris-catching solution is primarily intended for
RETROFIT OF WALLS
465
Figure 17.12 Example of the type of composite panels that may be used for retrofitting
masonry and stud walls to enhance their blast resistance
non-load-bearing walls whose function is not critical to preventing failure of the
structural system. However, these systems may also be applied to bearing walls,
where the breadth of the wall failure incurred can be spanned by the remaining
structural system.
In 2004, a series of reinforced and unreinforced polymer panels for retrofitting
masonry walls was developed (Crawford, Lan, and Wu 2009). These panels are
delivered to the site in the form of sheets, which are placed behind the wall and
anchored to the floors to hold them in place, as shown in Figure 17.11. These
panels have shown their worth against both vehicle and rocket blast threats in
tests in Israel and the U.S.
Retrofitting with Composite Panels In 2000, large-capacity catcher systems
were introduced and validated in blast tests (Crawford 2001) for retrofitting masonry or stud walls. These systems employed composite panels, like the one
shown in Figure 17.12, that were composed of FRP or steel sheets bonded to a
crushable core material, such as the rigid polyurethane foam shown in the figure.
The panel is placed behind the wall and anchored to the floor, generally using a
compliant anchorage. These systems have been proved successful in blast tests
for both masonry and metal stud walls, and were shown to work for very high
loads if the anchorage is designed appropriately (Crawford 2001, Crawford, Xin,
and Bogosian 2002).
Retrofitting with Conventional Methods Conventional methods (e.g., shotcrete
and dowels) may also be used to retrofit a masonry wall to enhance its blast
466
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
resistance, or just to prevent its debris from propagating into the building. These
systems should provide continuous support to the existing masonry wall to prevent localized breaching. Such systems would be expected to be less efficient
than those already mentioned. Moreover, the generally large quantities of materials used to install these types of retrofits may add to the risks from blast by
introducing considerably more material that could become a source of debris.
17.3.2 Stud Walls
Knox et al. (2000), and Crawford and Morrill (2003a, 2003b) also studied the
retrofit of wood and steel stud walls, using composite materials, which provided
an excellent means for adding blast protection to an existing stud frame building.
In the Crawford and Morrill studies, both fabric and sheet metal composites, such
as the one shown in Figure 17.12, were deployed.
17.4 FLOORS
Considerations of the blast protection afforded by the floors of a building are
likely to involve three kinds of situations, which are related to the location of the
detonation:
r Case 1: Considerations for detonations outside the building, where the detonation is some distance away. Here, enhancing the resistance of the floor
may be unnecessary or may be accomplished using conventional strengthening techniques, mainly because the floor’s response is neither very pronounced nor complex. The actual distance in question depends on the type
of exterior wall, the floor’s properties, and other factors.
r Case 2: Considerations for exterior detonations near the building, with no
wall present (e.g., near a loading dock). Here, an upward shock could be
imparted to a floor above the detonation and downward motion imparted to
the floor below the detonation.
r Case 3: Interior detonation. Here, considerations for two situations are of
issue: the blast above, and the blast below a floor. For an interior detonation,
gas pressures may be of major concern. In this case, adding venting might
be part of the retrofit strategy.
Retrofitting floors to provide protection from blast is likely to be more problematic than retrofitting any other structural component, mainly because of the
large spans and loaded areas, their load resistance being primarily related to gravity (i.e., with limited resistance to uplift), and the possibility that the detonation
may be quite close (e.g., a bomb resting on the floor). Moreover, in contrast to
most other retrofit strategies, the weight of the floor retrofit can be an issue because of the floor’s extent.
FLOORS
467
The following are three retrofit strategies for addressing the three threat situations mentioned above.
r Case 1. When the bomb is outside, the walls are likely to knock the blast
down sufficiently so that the floor response may be ignored—especially
since it’s likely in such a case that the walls would be retrofitted first, and
even less load would reach the floor. Moreover, the main concern is the net
load acting on the floor, which may be negligible, even if the walls fail.
However, in situations where the facades above and below a floor have very
different blast resistance (e.g., ground floors that have no exterior walls), a
substantial net upward (or downward) load on a floor can be generated, and
the situation may require some form of mitigation.
r Case 2. If the bomb is relatively close to the outside of a building and opposite an opening in the wall (or possibly a glass facade), the floor, especially
above the bomb, which would be loaded in its weak direction, may be a
cause of concern. Here, however, the effects of gas pressure can probably
be ignored.
r Case 3. For an interior detonation, both shock and gas pressure components
of the blast may cause concern.
The risks associated with a floor hit by a blast depend greatly on the size and
location of the bomb. For example, is the detonation from a bomb on the floor,
or above/below the floor? This has mostly to do with proximity and whether the
response is mainly localized or involves the whole of the floor, the magnitude of
velocity (shock) imparted to the floor and possible debris generated from it, and
whether the floor is loaded on its strong or weak axis.
Protection for floors against nearby bombs (i.e., within a few feet) is likely to
be impractical. For example, for a bomb resting on a conventional floor, there are
few (or no) reasonably priced ways to prevent the floor’s breach. Where the blast
is above the floor, then attaching a catcher system to the joists to prevent debris
propagation could be a good strategy.
For situations involving a vented blast occurring beneath a floor that has
a reasonable standoff, retrofitting the floor with an FRP may suffice. However, even if the floor remains relatively undamaged, there is likely to be significant shock imparted to it, which can cause severe injuries to those standing on it and launch furnishings, providing another source for injuries. Better
protection can be provided by a blast shield, possibly hung below the actual
floor, to either mitigate or prevent the airblast from striking the floor. In this
circumstance, not only is the floor not damaged, but the velocity imparted
by the shock, resulting in the launch of furnishings and people, is avoided
altogether.
The selection of a retrofit strategy for floors is likely to primarily involve
limiting the risks, rather than eliminating them. Three basic retrofit strategies,
depending on the bomb’s size and location, seem most applicable to addressing
468
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
the vulnerability presented by floors:
r Strengthening the existing floor through the addition of ductility and
capacity.
r Debris containment. This strategy is particularly useful for those floors in
direct contact with or very near to the bomb.
r Shielding. Shielding a floor from the blast load avoids the issue altogether.
This could also include constructing a second floor directly above/below the
existing one that is directly exposed to the threat. Alternatively, the existing
floor may be retrofitted so that it can provide the protection for a new floor.
Retrofit strategies for floors are likely to be much more difficult to implement
than for walls, which would imply that the best strategy here is to prevent a bomb
or an airblast from getting close enough to present a serious risk to the floors,
and to use the walls of the building to protect the floors. Even for a loading dock,
using roll-up doors or some other form of closure can probably deflect the blast
enough so that a floor above it is sufficiently protected.
17.5 BEAMS/GIRDERS/CONNECTIONS
The retrofit strategies pertaining to enhancing the blast protection afforded by
structural components that act as beams, girders, and the connections associated
with them—hereafter collectively referred to as beams—are generally of less
concern than the components already mentioned, because in many ways these
components generally present less of a risk.
Retrofit strategies for beams to accommodate blast loads are likely to be relatively simple and probably similar to those that might be used in conventional
retrofit strategies, especially for blast loads that are generated by charges that
predominantly excite the fundamental modes of the beam.
In addition, whether the blast is pushing up or down on the beam can make
a substantial difference in its response. Many beams are not as well attached in
the upward direction, and may have similar weaknesses in terms of their ability
to resist lateral loads. For closer charges, retrofit strategies are likely to be sitespecific and not easily addressed with simplified tools and conventional retrofit
strategies.
Blast effects data for beams (Crawford, Magallanes, and Morrill 2005,
Magallanes and Morrill 2008), although sparse, indicate that the beams themselves are likely to be less of an issue than their support connections. For external
blast, it is likely that the net loads pushing up/down on the beam will be negligible. However, loading on a beam can be quite localized (e.g., a bomb placed on
the floor immediately above the beam), or broad-based; for example, if applied to
a robust floor system that is strongly coupled to the beams framing it, the beam’s
loading would derive from a large tributary area.
SUMMARY
469
A major concern for beams is their attachment (i.e., connections) to their support (e.g., columns). For many steel beams, especially in nonseismic designs,
this may consist of only a shear tab, or for precast beams only a small ledger to
which it is only lightly attached (e.g., the shear tab shown in Figure 17.7b).
Several retrofit concepts have been developed for seismic applications to improve the anchorage of beams/girders at their supports that would also work in
blast environments, although they may need to be more robust. Another retrofit
strategy that has been used (Crawford and Lan 2005) relies on cabling attached to
the beams to carry the load, should their connection to the columns be damaged.
17.6 STRUCTURAL SYSTEM
The preceding sections addressed ways to retrofit structural components to enhance their survivability and protective capability in the event of a blast. It is
also important, perhaps even more so, to evaluate the structural system’s performance, especially once the retrofits are in place.
However, evaluating the risks to the system and describing retrofit strategies
for addressing them is beyond the scope of this chapter. With regard to assessing risks to the system, however, it is important to recognize that retrofitting
structural components can produce deleterious effects in terms of the system’s
performance. For instance, there would be a need to ensure that the impact of
the additional loading imparted to the system by any enhancement in the protection be evaluated—for example, the added lateral load imparted to the system
by the addition of blast-resistant walls could exacerbate the lateral response of a
building system sufficiently to cause its failure from excessive story drift.
17.7 SUMMARY
Information was presented in this chapter pertaining to retrofitting a facility’s
structural framing system and components to enhance the protection afforded
by them in the event that a blast occurs. Each of the major components of the
structural system was addressed, including columns, walls, beams, and floors.
Several strategies were described for mitigating risks from blast for each type of
component.
17.7.1 Inexact Science
The ideas presented in this chapter concerning the retrofit of structural components to enhance the blast protection afforded by them can only, due to space
limitations, reflect a general notion of strategies and methods that may be pursued in this endeavor. Moreover, only a few of the ways this might be accomplished could be cited, and all the concerns and complexities that might arise in
an actual application could not be addressed. What is intended, however, is to
apprise those desirous of developing retrofit strategies and designs of some of
470
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
the main issues and methods that have been found useful and effective, and to
describe extant test data and studies that are available.
Developing retrofit designs for any structure is often more art than engineering because each situation is likely to be unique, to have issues of unintended
consequences, to raise difficulties of identifying the extant design, and so forth.
In addition, blending with the existing structure, architecture, and functionality
is likely to present a host of conflicting constraints, making the retrofit even more
difficult.
It is likely that any retrofit design will not be able to offer the same certainty
of function as a new design. Moreover, achieving too high a certainty is likely
to result in a poorer level of protection or perhaps none at all—that is, there is
a certain paralyzing effect that can easily be reached by demanding too much
certainty. On the other hand, there is a particular need in this type of engineering
to be certain that the design is not fatally flawed. Adopting a good enough standard may offer the best course for this type of work, the main issue here being
the difficulty of defining such a standard ahead of time (e.g., for the purposes of
generating a contract and design budget).
17.7.2 Complexities
Retrofitting a structure to enhance its resistance to blast may require an inordinate
level of skill and experience. The limited amount of information presented in this
chapter can only allude to some of the considerations that are needed to develop
these retrofits and their integration with the existing structure, and should be
viewed as just an introduction to the subject.
It is important to recognize that, in many cases, the intensity and rapidity of
the blast load generate behaviors substantially different than generally encountered in structural design problems. Moreover, the forces and material damage
incurred are often difficult to determine without using HFPB finite element models, and for most retrofit designs, they are likely to be much higher than ordinarily
dealt with, requiring a familiarity with material and dynamic response behaviors
in this high loading rate regime.
The level of complexity and difficulty encountered in developing retrofit designs is likely to be highly exacerbated when the blast load is from a large,
close charge. These situations can be addressed, but they require considerations
in addition to those that can be alluded to here. Moreover, in these situations,
many simplified engineering models commonly used for blast effects analysis
are likely to be deficient and misleading in their prediction of response.
REFERENCES
American Concrete Institute, Committee 440. 2002. Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures (ACI 440.2R-02).
Farmington Hills, MI: American Concrete Institute.
SUMMARY
471
Baylot J. T., B. Bullock, T. R. Slawson, and S. C. Woodson. 2005. Blast response
of lightly attached concrete masonry unit walls. Journal of Structural Engineering
131 (8): 1186–1193.
Baylot J. T., S. T. Dennis, and S. C. Woodson. 2000. Response of 1/4 scale concrete masonry unit (CMU) walls to blast. In Emerging Technologies in Fluids, Structures, and
Fluid/Structure Interactions, Vol. 1. Proceedings of the 2000 ASME Pressure Vessels
and Piping Conference, Seattle, WA, July 23–27, 2000. New York: American Society
of Mechanical Engineers.
Carney, P. and J. J. Myers. 2003. Shear and flexural strengthening of masonry infill walls
with FRP for extreme out-of-plane loading. Proceedings, Architectural Engineering
Institute Annual Meeting, Austin, TX, 2003.
. 2005. Out-of-plane static and blast resistance of unreinforced masonry
wall connections strengthened with FRP. In 7th International Symposium on FiberReinforced Polymer Reinforcement for Concrete Structures, ACI Symposium Publication 230, 229–248. Farmington Hills, MI: American Concrete Institute.
Conley C. H. and P. F. Mlakar. 1999. Blast resistance of CFRP reinforced masonry. Presented at the 70th Shock and Vibration Symposium, Albuquerque, NM, November
15–19, 1999.
Crawford J. E. 2001. Addressing workplace protection issues more effectively. Proceedings of the Society of American Military Engineers Symposium on Comprehensive
Force Protection, Charlestown, SC, November 1–2, 2001.
Crawford J. E., D. D. Bogosian, and J. W. Wesevich. 1997. Evaluation of the effects of
explosive loads on masonry walls and an assessment of retrofit techniques for increasing their strength. Presented at the 8th International Symposium on the Interaction of
the Effects of Munitions with Structures, Defense Special Weapons Agency, McLean,
VA, April 1997.
Crawford J. E. and S. Lan. 2005. Design and implementation of protection technologies for improving blast resistance of buildings. Presented at the Enhancing Building
Security Seminar, Singapore, March 2005.
Crawford J. E., S. Lan, and Y. Wu. 2009. Design methods for enhancing the blast resistance of masonry walls and windows with bonded and unbonded polymer sheets.
Proceedings of the 8th International Conference on Shock and Impact Loads on Structures, Adelaide, Australia, December 2–4, 2009.
Crawford J. E., J. J. Magallanes, and K. B. Morrill. 2005. Vulnerability of steel framing
systems. Proceedings of the 12th International Symposium on the Effects of Munitions
with Structures, New Orleans, LA, September 26–30, 2005.
Crawford J. E., L. J. Malvar, B. W. Dunn, and D. J. Gee. 1996. Retrofit of reinforced
concrete columns using composite wraps to resist blast effects. Proceedings of the
27th Department of Defense Explosive Safety Seminar, Las Vegas, NV, August 19–22,
1966.
Crawford J. E., L. J. Malvar, J. M. Ferritto, K. B. Morrill, B. W. Dunn, and P. Bong.
2003. Description of Design Software for Retrofitting Reinforced Concrete Columns
to Improve Their Resistance to Blast. (Report TR-01-16.2). Burbank, CA: Karagozian
and Case. Limited Distribution.
Crawford J. E., L. J. Malvar, K. B. Morrill, and J. M. Ferritto. 2001. Composite retrofits to
increase the blast resistance of reinforced concrete buildings. Proceedings of the 10th
472
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
International Symposium on Interaction of the Effects of Munitions with Structures,
San Diego, CA, May 2001.
Crawford, J. E. and K. B. Morrill. 2002. New highly ductile designs for preventing injurious wall debris due to a terrorist bombing. Proceedings of the Protection of Structures
Against Hazards, Singapore, November 2002.
. 2003a. Retrofit techniques using polymers and fiber reinforced polymers for
preventing injurious wall debris. Proceedings of the 6th International Symposium on
Fibre-Reinforced Polymer Reinforcement for Concrete Structures, FRPRCS-6, Singapore, July 8–10, 2003, pp. 1199–1208.
. 2003b. Retrofit techniques for strengthening stud walls subjected to nearby
blast loads. Proceedings, 74th Shock and Vibration Symposium, San Diego, CA, October 27–31, 2003. Limited Distribution.
Crawford J. E., J. Wesevich, J. Valancius, and A. Reynolds. 1995. Evaluation of jacketed columns as a means to improve the resistance of conventional structures to blast
effects. Proceedings, 66th Shock and Vibration Symposium, Biloxi, MS, October
30–November 2, 1995.
Crawford J. E., X. Xin, and D. D. Bogosian. 2002. Retrofit techniques for conventional
walls to prevent their causing injurious debris due to a terrorist bombing. Proceedings
of the 17th International Symposium on Military Aspects of Blast and Shock, Las
Vegas, NV, June 10–14, 2002.
Davidson J. S., J. W. Fisher, M. I. Hammons, J. R. Porter, and R. J. Dinan. 2005. Failure
mechanisms of polymer reinforced concrete masonry walls subjected to blast. Journal
of Structural Engineering 131 (8): 1194–1205.
Davidson J. S., J. R. Porter, R. J. Dinan, M. I. Hammons, and J. D. Connell. 2004. Explosive testing of polymer retrofit masonry walls. Journal of Performance of Constructed
Facilities 18 (2): 100–106.
Defense Special Weapons Agency (DSWA). 1998. Design and Analysis of Hardened
Structures to Conventional Weapons Effects (TM 5-855-1, AFPAM 32-1147(1), NAVFAC P-1080, DAHSCWEMAN-97). Washington, DC: The Departments of the Army,
Air Force, Navy, and The Defense Special Weapons Agency.
Dennis S. T., J. T. Baylot, and S. C. Woodson. 2002. Response of 1/4 scale concrete masonry unit (CMU) walls to blast. Journal of Engineering Mechanics 128 (2):
134–142.
Federal Emergency Management Agency, Building Performance Analysis Team. 1996.
The Oklahoma City Bombing: Improving Building Performance through MultiHazard Mitigation (FEMA 277). Washington, DC: Federal Emergency Management
Agency.
Hammons M. I., K. J Knox, and J. R. Porter. 2002. Polymer materials for blast retrofit
of masonry construction. Proceedings, 30th Department of Defense Explosives Safety
Seminar, Atlanta, GA, August 13–15, 2002.
Johnson C. F. and T. R. Slawson. 2004. Concrete masonry units walls retrofitted
with fiber reinforced elastomeric systems for blast loads. Presented at the 75th
Shock and Vibration Symposium, Virginia Beach, VA, October 27–31, 2004. Limited
Distribution.
Johnson C. F., T. R. Slawson, T. K. Cummins, M. B. Hueste, and L. Beason. 2003.
Concrete masonry units (CMU) static and dynamic wall experiments with elastomeric
SUMMARY
473
retrofits. Presented at the 74th Shock and Vibration Symposium, San Diego, CA, October 27–31, 2003. Limited Distribution.
Johnson C. F., T. R. Slawson, T. K. Cummins, and J. L. Davis. 2004. Concrete masonry
unit walls retrofitted with elastomeric systems for blast loads. Presented at the 24th
Army Science Conference, Orlando, FL, November 29–December 2, 2004.
Kersul A. M. and D. Sunshine. 2002. Structural collapse blast mitigation. Presented at
the 17th International Symposium on Military Aspects of Blast and Shock, Las Vegas,
NV, June 10–14, 2002.
Knox K. J., M. I. Hammonds, J. R. Porter, and T. T. Lewis. 2000. Polymer materials for
structural retrofit. Proceedings, 29th Department of Defense Explosives Safety Seminar, New Orleans, LA, July 18–20, 2000.
Magallanes J. M. and K. B. Morrill. 2008. Fast running model for the residual capacity
of steel columns damaged by blast & fragment loads. Proceedings of the 79th Shock
and Vibration Symposium, Orlando, FL, October 26–30, 2008.
Malvar L. J. 1998a. Durability of composites in concrete. Presented at First International Conference on Durability of Composites for Construction, Sherbrooke, Quebec,
Canada, August 5–7, 1988, pp. 361–372.
. 1998b. Review of static and dynamic properties of steel reinforcing bars. ACI
Materials Journal 95 (5): 609–616.
Malvar L. J., J. E. Crawford, and K. B. Morrill. 2007. Use of composites to resist blast.
Journal of Composites for Construction 11 (6): 601–610.
Malvar L. J., K. B. Morrill, and J. E. Crawford. 1999. CTS-1 Retrofit Designs: For Reducing the Vulnerability of an Office Building to Airblast (Report TR-98-39.2). Burbank,
CA: Karagozian & Case.
. 2004. Numerical modeling of concrete confined by fiber reinforced composites. Journal of Composites for Construction 8 (4): 315–322.
Malvar L. J. and C. A. Ross. 1998. Review of static and dynamic properties of concrete
in tension. ACI Materials Journal 95 (6): 735–739.
Malvar L. J. and J. W. Wesevich. 1996. Peak Compression Membrane Response Enhancements for One-Way and Two-Way Slab Resistance Functions (TM-96-1.2). Burbank,
CA: Karagozian & Case.
Morrill, K. B., J. E. Crawford, and J. M. Magallanes. 2007. “Development of Simplified Tools to Predict the Blast Response of Steel Beam-Column Connections,” Proceedings of the 2007 ASCE/SEI Structures Congress, Long Beach, CA, May 16–19,
2007.
Morrill K. B., L. J. Malvar, and J. E. Crawford. 1999. Retrofit design procedure for existing RC buildings to increase their resistance to terrorist bombs. Proceedings of the 9th
International Symposium on Interaction of the Effects of Munitions with Structures
(IEMS), Berlin, Germany. May 3–7, 1999.
Morrill K. B., L. J. Malvar, J. E. Crawford, and S. W. Attaway. 2000. RC column and
slab retro-fits to survive blast loads. Proceedings of the ASCE Structures Congress,
Philadelphia, PA, May 8–10, 2000.
Morrill K. B., L. J. Malvar, J. E. Crawford, and J. M. Ferritto. 2004. Blast resistant design
and retrofit of reinforced concrete columns and walls. Presented at ASCE Structures
Congress, Nashville, TN, May 23–25, 2004.
474
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Morrill K. B., L. J. Malvar, J. E. Crawford, G. Hegemier, and F. Seible. 2001. Fullscale testing of reinforced concrete column retrofits to resist blast loads. Proceedings
of the 10th International Symposium on Interaction of the Effects of Munitions with
Structures, San Diego, CA, May 2001.
R
Morrill K. B., K. Nicolaisen, and T. Hutchinson. 2005. Laboratory Testing of Kevlar
©
Reinforced Polyurea (Report TR-04-52.1). Burbank, CA: Karagozian & Case.
Muszynski L. C. 1998. Explosive field tests to evaluate composite reinforcement of concrete and masonry walls. In Fiber Composites in Infrastructure, Proceedings of the
Second International Conference on Composites in Infrastructure, Tucson, Arizona,
January 5–7, 1998, pp. 277–284.
Muszynski L. C. and M. R. Purcell. 2003. Use of composite reinforcement to strengthen
concrete and air-entrained concrete masonry walls against air blast. Journal of Composites for Construction 7 (2): 98–108.
Myers J. J., A. Belarbi, and K. A. El-Domiaty. 2003. Blast resistance of un-reinforced
masonry walls retrofitted with fiber reinforced polymers. Proceedings for The North
American Masonry Society 2003 Annual Meeting, Clemson, South Carolina, pp.
312–323.
. 2004. Blast resistance of FRP retrofitted un-reinforced masonry (URM) walls
with and without arching action. The Masonry Society Journal 22 (1): 9–26.
Patoary M. K. H. and K. H. Tan. 2003. Blast resistance of prototype in-built masonry
walls strengthened with FRP systems. Proceedings of the 6th International Symposium
on Fibre-Reinforced Polymer Reinforcement for Concrete Structures, FRPRCS-6, Singapore, July 8–10, pp. 1189–1198.
Perea, A. and K. B. Morrill. 2002. Component high explosive testing of bare and
retrofitted columns. Presented at the 17th International Symposium on Military Aspects of Blast and Shock, Las Vegas, NV, June 10–14, 2002.
.Piepenburg, D., R. Martinez, R. Del Frate, and K. Morrill. 2002. Blast Resistance Upgrade for Glazing, Window Frame, and Wall Systems—Non-Load Bearing In-Fill
Walls with Blast Retrofits applied to the Wall, Window Frame and Glaze (Report No.
SAIC-03/1001). San Diego, CA: Science Applications International Corporation.
. 2003. Blast Resistance Upgrade for Glazing, Window Frame, and Wall Systems – Load Bearing Walls with Blast Retrofits applied to the Wall, Window Frame and
Glaze (Report No. SAIC-03/1003). San Diego, CA: Science Applications International
Corporation.
Simmons, L. 2005. Arena fragment experiments using mortars. Presented at the 76th
Shock and Vibration Symposium, Destin, FL, October 30–December 4, 2005.
Stanley, M., R. Martinez, and J. Koenig. 2005a. UL-Like Testing of Commercial Off-theShelf Products that Enhance the Blast and Ballistic Resistance of Structures, Quick
Look Report 6 (New Mexico Tech, Energetic Materials Research and Testing Center,
TR-05-08). Burbank, CA: Karagozian & Case. Limited Distribution.
. 2005b. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 7 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-05-14). Burbank, CA:
Karagozian & Case. Limited Distribution.
Stanley, M., J. Metzger, R. Martinez, and J. Koenig. 2005. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance the Blast and Ballistic Resistance of
SUMMARY
475
Structures, Quick Look Report 5 (New Mexico Tech, Energetic Materials Research and
Testing Center, TR-05-03). Burbank, CA: Karagozian & Case. Limited Distribution.
Stanley, M., J. Metzger, K. B. Morrill, and R. Martinez. 2005a. UL-Like Testing of
Commercial Off-the-Shelf Products that Enhance the Blast and Ballistic Resistance of
Structures, Quick Look Report 2 (New Mexico Tech, Energetic Materials Research and
Testing Center, TR-04-48). Burbank, CA: Karagozian & Case. Limited Distribution.
. 2005b. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 3 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-04-50). Burbank, CA:
Karagozian & Case. Limited Distribution.
. 2005c. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 4 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-04-55). Burbank, CA:
Karagozian & Case. Limited Distribution.
Tan K. H. and M. K. H. Patoary. 2004. Strengthening of masonry walls against outof-plane loads using fiber-reinforced polymer reinforcement. Journal of Composites
for Construction 8 (1): 79–87.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
INDEX
A
acetone peroxide, 26
Alfred P. Murrah Federal Building, 17,
22, 27, 173, 446, 447
alternative load path, 8
Amman, 25
analysis methods, 243, 244, 256, 389
Baker-Strehlow-Tang (BST), 179
blast curves, 179
component modeling, 249
computational fluid dynamics (CFD),
178
connections, 252
effects of preload, 251
envelope design approaches, 265–67
equivalent static loads, 112
explicit dynamic finite elements,
256
finite elements, 120, 245, 269
for moment frames, 246
for steel-braced frames, 247
multi-degree-of-freedom, 254, 256,
268
P-I diagrams, 253, 267
precast tilt-up construction, 247
precision of, 15
ray-tracing procedures, 209
reinforced masonry, 247
resistance factors, 112
semi-empirical, 210
shear walls, 247
simplified, 4, 251
single-degree-of-freedom, 9, 95, 123,
245, 254, 268
sophisticated approaches, 9
TNO Multi-energy Method (MEM),
179
antiterrorism, 102
appurtenances, 287
assessment
asset value, 18, 31
asset value ratings, 34
design-base threat, 51
interior explosion risk, 298
naturally occurring explosions, 58
risk, 4, 28, 38, 61, 242, 243
risk assessment models, 20
site-specific risk, 18
threat, 18, 51, 61
threat mapping, 36
threat ranking scale, 31
threat rating, 31
vulnerability, 18, 34, 54
vulnerability checklist, 37
vulnerability rating, 35, 37
Atlanta Centennial Olympic Park, 25
B
Baghdad, 23
base isolation, 292
of mechanical systems, 337
Beirut U.S. Embassy, 22
Beirut U.S. Marine headquarters, 22
blast consultants
qualifications, 14
blast design purposes, 89
blast design strategies, 88
blast environment, 5
blast loads, 4, 186
air-induced ground shock, 354
air-shock-induced motions, 356
angle of incidence, 189, 192
calculated by computational fluid
dynamics (CFD), 183
combining shocks from multiple
transmission media, 357
confidentiality, 13, 30
confined explosions, 199
477
478
INDEX
blast loads (Continued )
design base threat, 29, 52, 61, 71,
243
designations, 29
determination of, 55
direct-induced ground shock, 355
drag coefficients, 197
dynamic wind pressure, 185
empirical method of calculation,
183
factors of safety, 64
in reentrant corners, 293
in-structure shock, 353
leakage, 7, 207, 350, 351
methods of prediction, 175
on overhangs, 293
ray-tracing procedures, 209
rear wall, 198
reflected, 187
roof, 197
side wall and roof, 193
sources for guidance, 62
verification, 269
blast parameters
prediction of, 172
surface bursts, 174
blast pressures
as a function of standoff, 310
clearing, 162
deflagration, 162, 163
detonation, 163
propagation through ducts, 352
rarefaction, 162
reflected, 162
blast response, compared to seismic
response, 12
blast simulators, 154, 155
blast walls, 12, 328
blast wave. See shock wave
blowout panels, 282
bombs, delivery mechanisms, 62
building codes, 79
building envelope, 46
balanced or capacity design, 267
connections, 272
convex vs. concave, 294
curtain walls, 276
design objective, 264
operable windows, 278
point and cable supported curtain
walls, 278
robustness, 53
building programming,
dispersal/distribution of, 54
business continuity, 47
C
catch systems, 7, 287, 441, 464
chemical, biological, and radiological
(CBR) control, 48
collateral damage, 45
components
blast walls, 12
frangible, 6
non-structural, 9
sacrificial elements, 10
shielding, 10
concrete
beam detailing, 376
beam reinforcement, 377
beam splice locations, 378
beam-column joint detailing, 378
column detailing, 373
column reinforcement, 374
column retrofit, 453
column splice locations, 376
composite, 291, 388
confinement, 455
detailing, 371
detailing for direct shear, 375, 378
detailing for shear strength, 379
direct shear, 455
direct shear failure, 143
dynamic design stresses, 139
dynamic increase factor, 137
effect of strain rate on compressive
strength, 131, 140
effect of strain rate on tensile strength,
132
exposed columns, 288
failure modes, 141, 369
flat slabs, 243, 290
flexure, 455
mechanical properties for design,
138
plain, 134
punching shear failure, 143
reinforcement, 133, 136, 378
INDEX
reinforcing steel splices, 372
scabbing, 144, 249
shear, 454
slab connection detailing, 380
slab detailing, 379
slab reinforcement, 380, 389
spalling, 144, 249
strength reduction factors, 145
stress-strain relationships, 129, 130
stress-strain relationships in tension,
131
wall connection detailing, 381
wall detailing, 381
wall reinforcement, 381
concrete reinforcement, stress-strain
relationships, 132
confidentiality, 13, 30, 70
consequence evaluation, 44
consequence management, 18, 42, 57
consequence mitigation, 45
constitutive models, reinforced concrete,
132
consultants
qualifications, 14
teaming, 18, 31, 45, 49, 52, 56, 59, 68,
73, 266
control centers, security, 76
counterterrorism, 27
courtyards, 278
Crime Prevention through Environmental
Design (CPTED), 76
critical assets, 33
critical functions
dispersal, 18
D
damage expectations, based on levels of
protection, 90
damage level for petrochemical facilities,
106
Dar es Salaam, 22
defense in depth, 32, 75
deflagration. See blast pressures
deflection limits, 87
concrete components, 108
concrete members, 93
doors, 109
ductility ratio, 91, 94, 95, 96, 101, 103,
107, 108, 109, 113, 114, 393
479
high or heavy damage, 114
limitations on, 114
masonry components, 108
medium damage, 113
steel components, 107
steel members, 93
support rotation, 91, 93, 94, 95, 96,
101, 103, 107, 108, 109, 113, 114,
393
design approaches, 18, 424
balanced or capacity design, 271
design basis threat, 18, 29, 58
design philosophy
design to budget, 73
design to threat, 71
detonation. See blast pressures
disproportionate collapse. See
progressive collapse
doors, 107, 279, 281
ductility, 244
ductility ratio, 93, 126, 386, 388, 390,
396, 397, 407, See also deflection
limits
dynamic design stresses, 126
dynamic increase factor, 112, 124
concrete, 137
steel, 125
E
East Beirut U.S. Embassy annex, 22
egress, 12, 265
stairwell enclosures, 299
electrical system, 68
emergency response, 12, 44, 46, 48
evacuation, rescue, and recovery systems,
30, 48, 79, 265
emergency elevator system, 348
emergency lighting, 48
evaluation of, 54
fire, 49
fire alarm, detection, exiting, and
annunciation, 48
fire suppression, 48
redundancy, 49, 55
survivability, 49
video surveillance, 48
explosions
accidental, 58, 59
below grade effects, 290
480
INDEX
explosions (Continued )
boiling liquid expanding vapor
explosion (BLEVE), 170
bursting pressure vessels, 178
causes, 57
confined, 165
deflagration, 184
intentional, 60
interior, 174, 175–77, 298
leakage, 264
pressure vessel burst, 169
scaled ranges, 249
surface bursts, 174
vapor cloud, 168–69, 172, 178–80
venting of internal explosions, 175
explosive events, 19, 21, 248
management, 54
explosives
acetone peroxide, 26
ammonium nitrate and fuel oil
(ANFO), 164
C-4, 26, 59
high, 165, 172
low, 165
nitromethane, 165
Propellants and pyrotechnics, 165
RDX, 26, 59
Semtex, 26, 59
TNT, 26, 59, 163
triacetone triperoxide, 26, 27
F
facade. See building envelope
facility function redundancy, 55, 56
facility siting, 51
facility staffing and operations, 341
failure modes, 127
breach and spall, 433
diagonal tension, 370
direct shear, 249, 370, 433
flexure, 128, 141, 369, 429
flexure and axial load, 128, 144
fracture, 252
instability, 128, 371
lateral-torsional buckling, 392
local buckling, 392
membrane action, 370
P-Delta effects, 251
punching shear, 243
scabbing, 144
shear, 128, 142, 432
spalling, 144
steel columns, 457
fenestration, 273
fiber-reinforced polymer, 442
fire, 81, 163, 302, 305, 345
control center, 348
detection and alarm systems, 349
smoke-control systems, 349
flammable agents, 81
fragmentation, 248, 264, 329
armor-piercing fragment, 235
catch systems, 287
constrained fragments, 224
damage to roofs, 237
damage to various materials, 238
debris missiles, 7
design fragment parameters, 227
from pressure vessel bursts, 170
Gurney energy, 219
impact damage, 229
K constant, 224
K1 and K2 constants, 226
K3 penetration factor, 236
loadings, 216
material toughness, 226
non-armor-piercing fragment, 236
of glass, 46
penetration, 230
penetration into concrete targets,
234
perforation, 236
primary, 216, 217
range prediction, 229
secondary, 217, 219
spalling, 237
steel targets, 232
structural missiles, 10, 50
THOR constants, 231
THOR equation, 230
THOR equation variables for various
target materials, 233
trajectory, 228
unconstrained fragments, 220
velocity, 219, 227
frangible components, 6
Friedlander equations, 248
functional redundancy, 18, 48
INDEX
G
Glasgow International Airport, 24
glass, 274
analysis software, 275
H
Hopkinson scaled pressures and
impulses, 153
HVAC systems, 46
hydrocode, 175
I
improvised explosive devices (IED), 22
person-borne, 24
personnel, 61
vehicle-borne, 22, 30, 61
information sensitivity, 13
J
Jakarta JW Marriott Hotel, 23
K
Karachi Marriott Hotel, 24
Karachi U.S. Consulate, 23
kinetic energy, 10
L
large deformations, 9
levels of protection, 71, 90
load combinations, 392
load factors, 9
London, 27
louvers, 282
M
Madrid, 24, 25, 27
masonry
breach and spall, 433
damage expectations, 430
detailing, 435
effect of coursing, 426
failure modes, 427
grout, 426
mortar, 426
performance assumptions, 423
reinforcement, 425, 436
reinforcing steel splice capacity,
426
retrofit recommendations, 441
481
splice locations, 426
strength, 425
support connections, 439
unreinforced, 424, 440
wall analysis, 439
wall detailing, 438
mass, effect on blast response, 10
materials, constitutive relationships,
247
mechanical systems, 68
anchorage, 338, 359
design philosophy, 334
electrical systems, 345
emergency elevator system, 348
emergency power systems, 347
fire protection systems, 345
general announcing communication
systems, 349
HVAC, 46
HVAC and plumbing systems,
342
in hardened spaces, 342
level of protection, 335
lighting systems, 347
location of, 335, 341
plenums, 299
protection of, 265
segregation, 76
seismic vs. blast, 333, 336
shock induced by the structure,
336
shock isolation, 337, 339
system fragility, 337
testing, 333
mitigation strategies, 19, 60
momentum, 10
N
Nairobi, 22
O
occupant injuries, 8
operational security, 65
P
Paddy’s Bar, 23
peer review, 69, 157
Pentagon, 17
penthouses, 291
482
INDEX
performance criteria, 88, 386, See also
deflection limits
damage level, 97, 98
development bases, 99
explosive safety, 99, 101
facades, 270
for antiterrorism, 102, 103, 105
for petrochemical facilities, 105
publications, 100
performance goals, 87, 89
performance verification, 149
by analysis, 156
by testing, 150
perimeter protection, 310
active wedge, 321, 323
anti-ram systems, 33, 53, 317
beam barriers, 321, 323
berms and ditches, 325, 327, 328
bollards, 320, 322
cable-based systems, 324
costs, 312
crash modeling, 318
crash testing, 317
Crime Prevention through
Environmental Design, 33
pedestrian control barriers, 326
planters and surface barriers, 325,
328
personnel screening, 75
petrochemical facilities, 105
physical security, 65
planning, secure facility, 20
plenums, 298
plumbing system, 68
polyurea-like polymers
reinforced, 464
unreinforced, 463
polyvinyl butyral interlayer, 280, 299
post-explosion conditions, 77
potential energy, 11
pressure-impulse diagram, 95, 96
progressive collapse, 11, 46, 53, 67, 77,
257, 285, 303, 384, 385
alternative load path, 12
European guidance, 259
U.S. guidance, 259
protection category, descriptions,
102
pyrotechnic reaction, 168
Q
quality assurance, 69
R
recovery and contingency planning,
56
redundancy
of assets, 75
of building systems, 34
reinforcing steel, 136
effect of strain rate on tensile strength,
134, 140
effect of strain rate on yield strength,
133, 140, 424
splice capacity, 426
strain-rate effects, 136
stress-strain relationships, 133
resistance-deflection curves, 92
resource allocation, 71, 73
response criteria. See deflection limits
response limits, 87
response parameters, 91
restoration of operations, 47
retrofit
analysis and design tools, 448
beams and connections, 469
column test data, 447
concrete columns, 447
fiber reinforced polymer, 449, 450,
452, 460, 461
floors, 467
masonry walls, 459
polyurea-like polymers, 463
steel columns, 455
steel jacket column, 449, 451
steel sections, 458
stud walls, 467
tests, 451, 461
walls, 459
with composite panels, 466
risk acceptance, 70, 74
risk assessment. See also assessment
risk assessment process, 19
risk management, 19, 40
strategies, 43
risk mitigation, 39
risk rating, 38
risk reduction, 41, 72
Riyadh, 23, 27
INDEX
roofs, 290
concrete, 290
gardens, 291
steel, 290
S
sacrificial elements, 10
safe havens, 300
access to egress, 305
design requirements, 302
location, 304
mechanical systems, 342
screening
of cargo and mail, 76
of personnel, 75
of vehicles, 76
security design guidelines, 82
security zone, 75
Shanksville, Pennsylvania, 17
shielding, 10
shock wave, 163, 170
comparison of free field and reflected
blast loads, 171
from detonation, 164
high explosives, 184
Hopkinson-Cranz scaling method, 173
overpressure, 171
parameters, 170
propagation, 162
scaling, 171
site-specific risk assessment, 18
siting of facilities, 36, 293, 314, 326
balancing hardening with standoff, 310
skylights, 279
stairwell enclosures, 299
standoff, 51, 53, 55, 64, 308
pressures as a function of, 310
standoff military ordnance
rocket-propelled grenade, 30
shoulder fired missiles, 30
steel
column retrofit, 455
composite, 291, 388
connection design, 419
connection ductility, 387
constitutive models, 120
deck, 290
design examples, 398
detailing, 394
483
detailing for failure modes, 394
ductility, 386
dynamic design stress, 391
dynamic increase factors, 125, 390,
391, 420
effect of strain rate on mild steel, 122
material properties, 389
overstrength factor, 387
strain hardening, 126
strain-rate effects, 124, 387
strength increase factors, 390
strength reduction factors, 145
stress-strain relationships, 119
tensile strength, 121
yield strength, 121
strain rates, 9
strength reduction factors, 8
concrete, 145
steel, 145
stress increase factor. See dynamic
increase factor for various materials
stress-strain relationships
concrete, 129
steel, 119, 120
structural behavior, 8, 256
large deformations, 252
P-Delta effects, 11, 251
seismic vs. blast, 242, 266
structural design
general guidance, 65
impacts on facility programming, 68
structure
braced frames, 243
exposed structural systems, 287, 388
influence on blast load, 6, 244
moment frames, 243, 246
support rotation. See deflection limits
T
Taba Hilton, 24
target attractiveness, 47
tensile strength
effect of strain rate on, 121
effect of temperature on, 121
tension membrane, 93
testing
building components, 151
mechanical systems, 333
vehicle barriers, 150
484
INDEX
threat
identification and rating, 18, 28, 30,
298
map, 56, 58
methods to respond, 39
mulit-hazard, 301
probability and consequence rating, 30
sequential events, 301
threat reduction, 57, 72
TNT equivalence, 165
of high explosives, 166
of vapor cloud explosions, 169
triacetone triperoxide, 26
trinitrotoluene. See explosives, TNT
U
Unabomber, 25
utility siting, 53
V
value engineering, 69
vapor cloud explosions. See explosions
vaults, 292
vehicle barrier testing, 150
vehicle screening, 76
vehicle speed, 315
verification, 269
vulnerability
assessment planning, 34
reduction of, 63
vulnerability assessment, 18
W
walls
blowout panels, 282
concrete, 283
exterior, 282
masonry, 286
metal panels, 287
post-tensioned panels, 286
precast panels, 284
progressive collapse resistance, 285
retrofit, 459
secondary, 442
shotcrete, 442
stud, 286, 467
tilt-up construction, 286
windows, 109, 273
anchorage, 440
Department of Defense hazard levels,
110
frames and anchorages, 280
Interagency Security Committee
hazard levels, 110, 111
mullions, 279
operable vs. inoperable, 278
supporting structure, 281
World Trade Center, 17, 22, 23
Y
yield stress
effect of temperature on, 121
effects of strain rate on, 121
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
HANDBOOK FOR
BLAST-RESISTANT
DESIGN OF BUILDINGS
HANDBOOK FOR
BLAST-RESISTANT
DESIGN OF BUILDINGS
Edited by
Donald O. Dusenberry
JOHN WILEY & SONS, INC.
To my wife, Alice
∞
This book is printed on acid-free paper.
C 2010 by John Wiley & Sons, Inc. All rights reserved.
Copyright
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise,
except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA
01923, (978) 750-8400, fax (978) 646-8600, or on the web at www.copyright.com. Requests to the
Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons,
Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at
www.wiley.com/go/permissions.
Limit of Liability/Disclaimer of Warranty: While the publisher and the author have used their best
efforts in preparing this book, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this book and specifically disclaim any implied
warranties of merchantability or fitness for a particular purpose. No warranty may be created or
extended by sales representatives or written sales materials. The advice and strategies contained
herein may not be suitable for your situation. You should consult with a professional where
appropriate. Neither the publisher nor the author shall be liable for any loss of profit or any other
commercial damages, including but not limited to special, incidental, consequential, or other
damages.
For general information about our other products and services, please contact our Customer Care
Department within the United States at (800) 762-2974, outside the United States at (317)
572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print
may not be available in electronic books. For more information about Wiley products, visit our
web site at www.wiley.com.
Library of Congress Cataloging-in-Publication Data:
Handbook of blast resistant design of buildings / edited by Donald O. Dusenberry.
p. cm.
Includes index.
ISBN 978-0-470-17054-0 (cloth)
1. Building, Bombproof. I. Dusenberry, Donald O.
TH1097.H36 2010
693.8 54–dc22
2009019203
ISBN: 978-0-470-17054-0
Printed in the United States of America
10 9 8 7 6 5 4 3 2 1
CONTENTS
Preface
xv
Contributors
xix
I
DESIGN CONSIDERATIONS
1 General Considerations for Blast-Resistant Design
Donald O. Dusenberry
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
1.10
Introduction
Design Approaches
The Blast Environment
Structure As an Influence on Blast Loads
Structural Response
Nonstructural Elements
Effect of Mass
Systems Approach
Information Sensitivity
Summary
References
2 Design Considerations
Robert Ducibella and James Cunningham
2.1
2.2
2.3
2.4
Introduction
A New Paradigm for Designing Blast-Resistant Buildings,
Venues, and Sites
A Brief History of Recent Terrorist Attacks
2.3.1 Terrorists’ Use of Explosives
2.3.2 Vehicle-Borne Improvised Explosive Devices
2.3.3 Person-Borne Improvised Explosive Devices
2.3.4 Locally Available Explosives
2.3.5 Some Counterterrorism Considerations
Collaborating to Analyze Risk
2.4.1 Step 1—Threat Identification and Rating
2.4.2 Step 2—The Asset Value Assessment
2.4.3 Step 3—The Vulnerability Assessment
1
3
3
4
5
6
8
9
10
12
13
14
15
17
17
18
21
21
22
24
25
27
28
28
31
34
v
vi
3
CONTENTS
2.4.4 Step 4—The Risk Assessment
2.4.5 Step 5—Considering Mitigation Options
2.4.6 The Continuing Role of Risk Management
2.5 Consequence Management
2.5.1 Consequence Evaluation
2.5.2 Function Redundancy
2.5.3 Building Location
2.5.4 Building Dispersal/Distribution of Functional Programs
2.5.5 Disaster Recovery and Contingency Planning
2.6 Threat Reduction
2.6.1 Accidental Explosions
2.6.2 Intentional Explosions
2.7 Vulnerability Reduction
2.7.1 Standoff Distance
2.7.2 Physical Security
2.7.3 Operational Security
2.7.4 Structural Design
2.8 Risk Acceptance
2.8.1 Design to Threat
2.8.2 Design to Budget
2.9 Some Recent Examples of Security Design “Best Practices”
2.10 Related Phenomena
2.10.1 Progressive Collapse
2.10.2 Disruption of Evacuation, Rescue, and
Recovery Systems
2.10.3 Attendant Fires
2.11 Security Design Consideration Guidelines
2.12 Conclusion
References
38
39
40
42
44
48
51
54
56
57
59
60
63
64
65
65
65
70
71
73
75
76
77
Performance Criteria for Blast-Resistant Structural Components
Charles J. Oswald
87
3.1
3.2
3.3
3.4
87
88
91
3.5
Introduction
Building and Component Performance Criteria
Response Parameters
Empirical Correlations between Response Parameters
and Component Damage
Response Criteria Development
3.5.1 Explosive Safety Criteria
3.5.2 Response Criteria for Antiterrorism
3.5.3 Response Criteria for Blast-Resistant Design of
Petrochemical Facilities
3.5.4 Blast Resistant Doors
3.5.5 Blast-Resistant Windows
79
81
83
84
85
95
99
99
102
105
107
109
CONTENTS
3.5.6 Response Criteria for Equivalent Static Loads
3.5.7 Comparisons of Published Response Criteria
3.6 Response Criteria Limitations
References
4 Materials Performance
Andrew Whittaker and John Abruzzo
4.1
4.2
Introduction
Structural Steel
4.2.1 Stress-Strain Relationships
4.2.2 Constitutive Models for Structural Steel
4.2.3 Component Level Strain Rate and Temperature Effects
4.2.4 Mechanical Properties for Design
4.2.5 Failure Modes of Structural Components
4.3 Reinforced Concrete
4.3.1 Stress-Strain Relationships for Concrete
4.3.2 Stress-Strain Relationships for Reinforcement
4.3.3 Constitutive Modeling of Concrete and Rebar
4.3.4 Component Level Strain-Rate Effects
4.3.5 Mechanical Properties for Design
4.3.6 Component-Level Failure Modes
4.4 Strength-Reduction Factors for Steel and Reinforced Concrete
References
5 Performance Verification
Curt Betts
5.1
5.2
5.3
Introduction
Performance Verification
Testing
5.3.1 Vehicle Barrier Testing
5.3.2 Building Components
5.4 Analysis
5.5 Peer Review
References
vii
112
113
114
116
119
119
119
119
120
123
125
127
129
129
132
132
136
138
141
144
145
149
149
149
150
150
151
156
157
157
II BLAST PHENOMENA AND LOADINGS
159
6 Blast Phenomena
Paul F. Mlakar and Darrell Barker
161
6.1
6.2
6.3
Introduction
Sources of Blasts
Characteristics of Blast Waves
161
162
170
viii
7
8
CONTENTS
6.3.1 Key Parameters
6.3.2 Scaling
6.4 Prediction of Blast Parameters
6.4.1 High Explosives
6.4.2 Bursting Pressure Vessels
6.4.3 Vapor Cloud Explosions
6.5 Summary
References
170
171
172
172
177
178
181
181
Blast Loading
Paul F. Mlakar and William Bounds
183
7.1 Introduction
7.2 Empirical Method
7.2.1 Empirical Method—Basic Blast Wave Example
7.3 Front Wall Loads
7.3.1 Empirical Method—Front Wall Loading Example
7.3.2 Empirical Method—Oblique Angle Example
7.4 Side Wall and Roof Loads
7.4.1 Empirical Method—Side Wall Loading Example
7.4.2 Empirical Method—Roof Loading Example
7.5 Rear Wall Loads
7.5.1 Empirical Method—Rear Wall Loading Example
7.6 Confined Explosions
7.7 Leakage
7.8 Ray-Tracing Procedures
7.9 Summary
References
183
183
186
186
188
192
192
194
196
197
197
198
206
208
212
212
Fragmentation
Kim King
215
8.1 Introduction
8.2 Debris
8.3 Loadings
8.3.1 Primary Fragmentation
8.3.2 Secondary Fragmentation
8.4 Design Fragment Parameters
8.4.1 Fragment Final Velocity
8.4.2 Fragment Trajectory
8.5 Fragment Impact Damage
8.5.1 Fragment Penetration into Miscellaneous Materials
(THOR Equation)
8.5.2 Steel
8.5.3 Fragment Penetration into Concrete Targets
8.5.4 Fragment Perforation of Concrete Targets
215
215
215
216
218
226
226
227
228
229
231
233
235
CONTENTS
8.5.5 Fragment Spalling of Concrete Targets
8.5.6 Roofing Materials
8.5.7 Other Materials
References
III SYSTEM ANALYSIS AND DESIGN
9
236
236
237
237
239
Structural Systems Design
Robert Smilowitz and Darren Tennant
241
9.1
241
241
243
244
245
246
248
251
252
252
253
255
255
256
258
258
261
9.2
9.3
9.4
10
ix
General Discussion
9.1.1 Seismic versus Blast
9.1.2 Analytical Methods
Modeling
9.2.1 Systems
9.2.2 Materials
9.2.3 Members
9.2.4 Connections
Analytical Approaches
9.3.1 P-I Diagrams
9.3.2 Single-Element Analyses
9.3.3 Structural Systems Response
9.3.4 Explicit Dynamic Finite Element Analyses
Progressive Collapse
9.4.1 European Guidance
9.4.2 U.S. Guidance
References
Building Envelope and Glazing
Eve Hinman and Christopher Arnold
263
10.1 Design Intent
10.1.1 Life Safety
10.1.2 Emergency Egress and Facilitating Search
and Rescue
10.1.3 Critical Functions (Protecting Equipment and
Business Processes)
10.2 Design Approach
10.2.1 Response Criteria
10.2.2 Static versus Dynamic
10.2.3 Balanced Design
10.2.4 Load Path
10.3 Fenestration
10.3.1 Glass
10.3.2 Mullions/Transoms
10.3.3 Frame and Anchorage
263
263
264
264
265
269
270
270
270
272
273
278
279
x
CONTENTS
10.4
10.5
10.6
10.7
11
12
10.3.4 Supporting Structure
10.3.5 Other Penetrations
Exterior Walls
10.4.1 Concrete Walls
10.4.2 Masonry
10.4.3 Steel
10.4.4 Other
Roof Systems
10.5.1 Concrete
10.5.2 Steel
10.5.3 Composite
10.5.4 Penthouses/Gardens
Below Grade
Reduction of Blast Pressures
References
280
280
281
282
285
285
286
289
289
289
290
290
290
292
294
Protection of Spaces
MeeLing Moy and Andrew Hart
297
11.1 Areas Isolating Interior Threats
11.2 Stairwell Enclosures
11.3 Hardened Plenums
11.4 Safe Havens
11.4.1 FEMA Documents
11.4.2 Multi-Hazard Threats
11.4.3 Design Requirements for Protective
Shelters
References
297
298
298
299
299
300
Defended Perimeter
Joseph L. Smith and Charles C. Ellison
307
12.1 Goals
12.2 Standoff
12.2.1 Balancing Hardening with Standoff
12.2.2 Balancing Costs
12.2.3 Site Planning
12.3 Vehicle Control Barriers
12.3.1 Crash Testing
12.3.2 Crash Modeling
12.3.3 Walls
12.3.4 Bollards
12.3.5 Active Wedge
12.3.6 Beam Barriers
12.3.7 Cable-Based Systems
307
307
309
311
313
316
316
317
319
319
320
320
323
301
305
CONTENTS
12.3.8
12.3.9
13
xi
Planter and Surface Barriers
Berms, Ditches, and Other Landscaping
Features
12.4 Pedestrian Control Barriers
12.5 Blast Walls and Berms
References
324
Blast-Resistant Design of Building Systems
Scott Campbell and James Ruggieri
331
13.1 Background
13.2 Introduction
13.3 Design Considerations
13.3.1 Level of Protection
13.3.2 Blast Pressures
13.3.3 Shock Induced by the Structure
13.3.4 Equipment/System Anchorage
13.3.5 Placement of Critical Systems Equipment and
Control Stations
13.3.6 Staffing and Building Operations
13.3.7 Construction of Hardened Spaces
13.3.8 HVAC and Plumbing Systems
13.3.9 Electrical Systems
13.3.10 Lighting Systems
13.3.11 Other Systems/Considerations
13.4 Loading Calculation
13.4.1 Blast Pressure
13.4.2 In-Structure Shock
13.5 Summary
References
331
332
333
334
334
335
337
324
325
327
329
340
340
341
341
344
346
346
348
349
352
362
363
IV BLAST-RESISTANT DETAILING
365
14
Blast-Resistant Design Concepts and Member Detailing
Steven Smith and W. Gene Corley
365
14.1 General
14.1.1 Scope
14.2 Failure Modes
14.2.1 Flexural
14.2.2 Diagonal Tension
14.2.3 Direct Shear
14.2.4 Membrane
14.2.5 Stability
367
367
368
368
369
369
369
370
xii
15
CONTENTS
14.3 Detailing
14.3.1
General
14.3.2
Splices
14.3.3
Columns
14.3.4
Beams
14.3.5 Beam-Column Joints
14.3.6
Slabs
14.3.7
Walls
References
370
370
371
372
375
377
378
380
380
Blast-Resistant Design Concepts and Member Detailing: Steel
Charles Carter
383
15.1 General
15.1.1 Typical Building Designs
15.1.2 Prescriptive Building Designs
15.1.3 Performance-Based Building Designs
15.2 Blast Effects on Structural Steel and Composite Structures
15.2.1 Member Ductility
15.2.2 Connection Ductility
15.2.3
Overstrength
15.2.4 Beneficial Strain-Rate Effects
15.2.5 Beneficial Effects of Composite Construction
15.2.6 Perimeter Column Design
15.2.7 Perimeter Girder Design
15.2.8
Slab Design
15.3 Analysis and Design of Structural Members
15.4 Steel Material Properties for Blast Design
15.4.1 Strength Increase Factor (SIF)
15.4.2 Dynamic Increase Factor (DIF)
15.4.3 Dynamic Design Stress
15.5 Design Criteria for Blast Design
15.5.1
General
15.5.2 Load Combinations
15.5.3 Resistance Factor and Factor of Safety
15.5.4 Local Buckling
15.5.5 Lateral-Torsional Buckling
15.5.6 Deformation Criteria
15.5.7 Detailing for Specific Failure Modes:
15.6 Examples
15.6.1 Example 1—Determining Capacities
15.6.2 Example 2—Design and Analysis for Blast Loads on
Members
15.7 Design of Connections
References
383
383
384
385
386
386
386
386
386
387
387
387
388
388
388
389
389
390
390
390
391
391
391
391
391
393
397
397
402
418
419
CONTENTS
16
17
xiii
Blast-Resistant Design Concepts and Member
Detailing: Masonry
Shalva Marjanishvili
421
16.1 General Considerations
16.1.1 Masonry
16.1.2 Reinforcement
16.1.3 Mortar
16.1.4 Grout
16.1.5 Construction Methods
16.2 Failure Modes
16.2.1 Flexure
16.2.2 Diagonal Tension Shear
16.2.3 Direct Shear
16.2.4 Breach and Spall Phenomena
16.3 Reinforced Masonry Detailing
16.3.1 General
16.3.2 Longitudinal Reinforcement
16.3.3 Horizontal Reinforcement
16.3.4 Walls
16.3.5 Support Connections
16.4 Unreinforced Masonry
16.4.1 Performance Evaluation
16.4.2 Retrofit Recommendations
References
423
424
424
425
425
425
426
428
431
432
432
434
435
435
435
438
438
439
439
440
442
Retrofit of Structural Components and Systems
John E. Crawford and L. Javier Malvar
445
17.1 Introduction
17.2 Retrofit of Columns
17.2.1 Reinforced Concrete Columns
17.2.2 Steel Columns
17.3 Retrofit of Walls
17.3.1 Masonry Walls
17.3.2 Stud Walls
17.4 Floors
17.5 Beams/Girders/Connections
17.6 Structural System
17.7 References
17.7.1 Inexact Science
17.7.2 Complexities
References
445
446
446
454
458
458
466
466
468
469
469
469
470
470
Index
477
PREFACE
The need for protection against the effects of explosions is not new. The use
of explosive weaponry by the military necessitated resistive entrenchments ages
ago. Industrialization of our societies well over a century ago meant that we
intended to manufacture, store, handle, and use explosives in constructive ways.
To support these military and industrial purposes, a relatively small group of
designers have worked to devise ways to strengthen the blast resistance of our
structures.
Early attempts at blast-resistance design necessarily relied on judgment, test,
and trial-and-error construction to find the best solutions. As technology improved, designers became better able to predict the influences of explosions
and the resistive responses that they strove to impart into their designs. More
recently, in the past several decades chemists, physicists, blast consultants, and
structural engineers have been empowered by technologies and computational
tools that have enhanced the precision of their analyses and the efficiency of their
designs.
At the same time, the need has increased. The small contingent of designers
skilled in the art and science of creating structural designs that will resist explosive forces has been joined by a larger group of architects, engineers, blast
consultants, and security consultants who are trying to respond to the increasing concern from a broader group of clients who fear an exposure that they did
not anticipate before and frequently did not bring upon themselves. Consultants
who have never before had to assess risks, devise risk-reduction programs, provide security systems, establish design-base threats, calculate the pressures and
impulses from explosions, and create cost-effective structural designs are being
thrust into the process. Many are ill-trained to respond.
There are several good references on some of the aspects of designing for
blast resistance. Some of these references support military purposes or for other
reasons have government-imposed restrictions against dissemination. As such,
they are not widely available to consultants working in the private sector. Nearly
all those references and the references that are public each treat an aspect of
blast phenomenology, security systems, and structural design for blast resistance,
but few, if any, bring together in one place discussions of the breadth of the
issues that are important for competent designs. Consultants are forced to collect
a library of references and extract from each the salient information that they
then synthesize into a comprehensive design approach.
xv
xvi
PREFACE
In addition, practitioners who do receive the limited-distribution references
for the first time or who find references that are public usually discover
immediately that designing for blast resistance is completely different from designing for any environmental load they encountered previously. Designers often
realize quickly that they are embarking on design process for which they do not
have the knowledge or experience for adequate competency. Those who do not
have this realization might be operating at risk if they are not careful and thorough students.
The purpose for this handbook is to bring together into one publication discussions of the broad range of issues that designers need to understand if they
are to provide competent, functional, and cost-efficient designs. The contributors
to this book are among the most knowledgeable and experienced consultants and
researchers in blast resistant design, and contribute their knowledge in a collaborative effort to create a comprehensive reference. Many of the contributors to
this handbook are collaborating in the development of the first-ever public-sector
standard for blast resistant design, being developed contemporaneously with this
handbook by the Structural Engineering Institute (SEI) of the American Society of Engineers. While there undoubtedly will be some differences between the
SEI standard and this handbook, many readers will consider these publications
as companions.
This handbook is organized into four parts, each addressing a range of aspects
of blast-resistance design.
Part 1: Design Considerations provides an overview of basic principles.
It has five chapters dealing with general considerations and the design process; risk analyses, reduction, and avoidance; criteria that establish acceptable performance; the science of materials performance under the extraordinary
blast environment; and performance verification for technologies and solution
methodologies.
Part 2: Blast Phenomena and Loadings, in three chapters, describes the
explosion environment, loading functions to be used for blast response analysis,
and fragmentation and associated methods for effects analyses.
Part 3: System Analysis and Design has five chapters that cover analysis and design considerations for structures. This part instructs on structural, building envelope, component space, site perimeter, and building system
designs.
Part 4: Blast-Resistant Detailing addresses detailing structural elements for
resistance. Chapters on concrete, steel, and masonry present guidance that is
generally applicable for new design. The fourth chapter addresses retrofits of
existing structures.
I wish to thank all the contributors for their commitment to this work, their
collaborative spirit, and, of course, their willingness to share the blast-related
expertise that they have presented in their chapters. I wish to thank Steven Smith
of CTLGroup in particular, for organizing and harmonizing the four chapters of
Part 4. William Zehrt of the Department of Defense Explosives Safety Board
improved the quality of this handbook by reviewing the chapters of Part 2.
PREFACE
xvii
I also wish to thank James Harper, Editor of John Wiley & Sons for supporting this effort; Daniel Magers, Senior Editorial Assistant, and Amy Odum
for her able supervision of the copyediting and production; and the copyeditors,
compositors, typesetters, and others of the publisher’s staff who have professionally assembled this book and brought it to publication.
Donald O. Dusenberry
Wakefield, Massachusetts
CONTRIBUTORS
John Abruzzo, P.E.
Principal
Thornton Tomasetti
555 12th Street, Suite 600
Oakland, CA 94607
Tel: (510) 433-9370
JAbruzzo@ThorntonTomasetti
.com
Christopher Arnold, FAIA, RIBA
President, Building Systems
Development Inc
Palo Alto CA
1248 Waverley Street
Palo Alto, CA 94301
Tel: 650-462-1812
[email protected]
[email protected]
Curt P. Betts, P.E.
Chief, Security Engineering Section
US Army Corps of Engineers
Protective Design Center
1616 Capitol Avenue
Omaha, Nebraska 68102-4901
Tel: (402) 995-2359
[email protected]
Darrell D. Barker, P.E.
Vice President
Extreme Loads and Structural
Risk
ABS Consulting
14607 San Pedro Ave., Suite 215
San Antonio, Texas 78232
Tel: (210) 495-5195
[email protected]
William Bounds, P.E.
Fluor
PO Box 5014
Sugar Land, Texas 77487
[email protected]
Scott Campbell, Ph.D., P.E.
Structural Analysis Consulting
Group
PO Box 91364
Louisville, KY 40291
Tel: (502) 762-9596
[email protected]
Charles Carter
American Institute of Steel
Construction
One East Wacker Drive
Suite 700
Chicago, Illinois 60601-1802
Tel: (312) 670-2400
[email protected]
W. Gene Corley, Ph. D.
CTLGroup
5400 Old Orchard Road
Skokie, Illinois 60077-4321
Tel: (847) 972-3060
Fax: (847) 965-6541
[email protected]
xix
xx
CONTRIBUTORS
John E. Crawford
Karagozian & Case
2550 N. Hollywood Way, Suite 500
Burbank, CA 91505-5026
Tel: (818) 240-1919
[email protected]
James D. Cunningham
Williamsburg, Virginia
Tel: (757) 645-4057
[email protected]
Robert Ducibella
Senior Principal
Ducibella Venter & Santore
Security Consulting Engineers
Sturbridge Commons – Franklin
House
250 State Street
North Haven, CT 06473
Tel: (203) 288-6490
[email protected]
Donald O. Dusenberry, P.E.
Senior Principal
Simpson Gumpertz & Heger Inc.
41 Seyon Street, Building 1, Suite 500
Waltham, MA 02453
Tel: (781) 907-9000
[email protected]
Chuck Ellison, P.E.
Senior Security Engineer
Applied Research Associates, Inc.
119 Monument Place
Vicksburg, MS 39180
Phone: (601)638-5401
[email protected]
Andrew Hart, Ph.D., MSc, BEng
(Hons), Aff.M.ASCE
Martinez, CA 94553
Tel: (925) 370-3866
[email protected]
Eve Hinman, Eng.Sc.D., P.E.
President
Hinman Consulting Engineers, Inc.
One Bush Street, Suite 510
San Francisco, CA 94104
Tel: (415) 621-4423
[email protected]
Kim W. King, P.E.
Director of Engineering
2195 Redwoods Crest
San Antonio, TX 78232
Tel: (210) 213-3737
[email protected]
L. Javier Malvar
Naval Facilities Engineering Service
Center
1100 23rd Avenue
Port Hueneme, CA 93043-4370
[email protected]
Shalva M. Marjanishvili, Ph.D.,
P.E., S.E.
Technical Director
Hinman Consulting Engineers, Inc.
One Bush Street, Suite 510
San Francisco, CA 94104
Tel: (415) 621-4423
[email protected]
Paul F. Mlakar, Ph.D., P.E.
U.S. Army Engineer Research and
Development Center
3909 Halls Ferry Road
Vicksburg, MS 39180
Tel: (601) 634-3251
[email protected]
MeeLing Moy, P.E.
President
The Link CE, PLLC
New York
Tel: (646) 385-5096
[email protected]
CONTRIBUTORS
Charles J. Oswald, Ph.D, P.E.
Protection Engineering Consultants
4203 Gardendale, Suite C112
San Antonio, TX 78229
Tel: (512) 380-1988
[email protected]
James Angelo Ruggieri, P.E.
10710 Timberidge Road
Fairfax Station, VA 22039
Tel: (703) 250-3671
[email protected]
Robert Smilowitz, Ph.D., P.E.
Weidlinger Associates, Inc.
375 Hudson Street
12th Floor
New York, New York 10014-3656
Tel: (212) 367-3090
[email protected]
Joseph L. Smith, PSP
Senior Vice President
Applied Research Associates, Inc.
119 Monument Place
Vicksburg, MS 39180
Tel: (601) 638-5401
[email protected]
Steven Smith
CTLGroup
10946 Eight Bells Lane
Columbia, MD 21044
Tel: (410) 997-0400
[email protected]
Darren Tennant
Weidlinger Associates, Inc.
6301 Indian School Road, NE,
Suite #501
Albuquerque, NM 87110
Tel: (505) 349-2820
[email protected]
Andrew Whittaker, Ph.D., S.E.
University at Buffalo
Buffalo, NY 14260
Tel: (716) 645-4364
[email protected]
xxi
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
I
Design Considerations
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
1
General Considerations for
Blast-Resistant Design
Donald O. Dusenberry
1.1 INTRODUCTION
Until recently, relatively few engineers and architects have had to design structures and their systems to resist the effects of explosions. Military engineering
personnel, consultants to the federal government, and consultants to industries
that use explosive or volatile materials constituted the primary population of designers routinely analyzing blast effects.
Following the explosion that demolished the Alfred P. Murrah Federal Building in Oklahoma City in 1995, members of the structural design and construction
industries have been increasingly quizzed by owners about blast-related hazards,
risks, and methods of protection. The types and numbers of clients seeking blast
resistance in their structures have expanded.
The terrorist events of the recent past and the fear that others may occur in the
future have led many businesses, particularly those with an international presence, to consider their vulnerability. And, of course, as their neighbors work to
enhance the performance of their buildings, owners and tenants who do not envision themselves as targets of malevolent acts nevertheless begin to wonder if
their structures might be damaged as a consequence of their proximity to targets. Some have argued that adding blast resistance and enforcing standoff for
one building on a block unfortunately increases the threat for others, because it
encourages aggressors to attempt to assemble bigger bombs and detonate them
closer to the target’s neighbors.
There seems to be a sense of anxiety about the vulnerability of our buildings,
bridges, tunnels, and utilities in the midst of numerous recognized international
social and political instabilities, and given the potential for domestic groups and
individuals to seek influence and create disruption by resorting to violent means.
As a result, consultants designing rather pedestrian buildings now are expected to
provide advice and sometimes specific enhancements in response to quantifiable
threats, as well as perceived vulnerabilities.
In this environment, engineers need training and information so that they can
provide designs that effectively enhance a building’s response to explosions.
3
4
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.2 DESIGN APPROACHES
Most engineers and architects serving clients with growing interest in blast resistance are uninitiated in the relevant design practice. Blast loading is very different from loadings commonly analyzed by structural engineers. Peak pressures
are orders of magnitude higher than those associated with environmental loads,
but their durations generally are extremely short compared to natural periods of
structures and structural components. In addition, given that the risk of an explosion at any one facility normally is very low and the costs to achieve elastic
response often are prohibitive, designs usually engage the energy-dissipating capability of structural and enclosure elements as they are deformed far into their
inelastic ranges. This forces engineers to account for geometric and material
nonlinearities.
At first, designing for blast resistance might sound similar to designing for
seismic resistance because neither is static and both rely on post-yield response.
But even those similarities are limited. The dominant frequencies of seismic excitations are on the order of the lowest natural frequencies of building response,
not much faster, as is generally the case for blast loadings. Blast loading usually
is impulsive, not simply dynamic.
While we tolerate some damage in earthquakes, to dissipate energy, we usually allow more damage for blast events. We expect facades to sustain severe
damage. In fact, blast-resistant design often tolerates breaching of the building
enclosure (with attendant risk of fatalities) and even sometimes partial collapse
of buildings.
Many blast-resistant designs require very sophisticated approaches for the
analysis of building response to explosions (National Research Council 1995).
There are techniques for accurate assessment of blast pressures and impulses
in complicated environments, modeling the influence of those blast loadings on
surfaces, and structural response to those loads. There are critical facilities and
blast conditions that warrant the use of these techniques. However, much blastresistant design is performed following simplified procedures (U.S. Department
of Defense 2008) that approximate actual conditions, and therefore lack high
fidelity. This often is appropriate because of, and at least in part follows from,
inevitable uncertainties that mask the phenomenon and the structure’s response.
In addition, there are practical matters of prudence, economics, and risk acceptance that drive analyses of blast response.
Risk analyses are important components of the design for blast resistance
(Federal Emergency Management Agency 2003). Among the products of such
analyses are estimates of the threat for which a structure should be designed. The
magnitude of intentional, nonmalevolent explosions and industrial explosions
sometimes can be estimated with precision commensurate with that of other
common loadings (Center for Chemical Process Safety 1996). The quantity of
explosive materials can be estimated, the potential locations of the design-base
explosion can be isolated, and often there are relatively few nearby objects that
significantly affect the shock front advance.
THE BLAST ENVIRONMENT
5
This is not the case for many accidental explosions and most malevolent explosions. The assessment of the threat in these instances often does not have a
probabilistic base. When sufficient data do not exist, consultants are forced to
use judgment rather than hard science to establish the threat.
When data are not available, consultants often establish the magnitude of the
threat of a malevolent explosion by assessing the probable size of the container
(e.g., letter, satchel, package) in which a bomb is likely to be delivered (U.S.
Department of Defense 2002a), and then selecting a design-base explosive mass
based on a fairly arbitrary assignment of the quantity of explosive that could
reasonably be accommodated in that container. In these cases, there is relatively
high uncertainty about the intensity of the explosion that might actually occur.
Obviously under these circumstances, there is a commensurate level of uncertainty about the outcome.
1.3 THE BLAST ENVIRONMENT
Engineers skilled in the design of buildings for occupancy-related and environmental loads (e.g., dead, live, wind, snow, and seismic loads), but faced with a
new challenge to design for blast loads, often find themselves ill-equipped for
the challenge. Designers are used to treating all other common loadings as either
static or quasi-static, because the rise time and duration for the equivalent load
are on the order of, or longer than, the longest natural periods of the structure.
Designing for blast loading generally cannot follow this approach.
Conventional design for common time-varying loads, including wind and
seismic, includes techniques that allow conversion of these dynamic phenomena
into quasi-static events that recognize and simplify the dynamics. Wind loads
defined in one of the most common references (American Society of Civil Engineers 2005) are based on an acknowledgment of the range of natural frequencies
of common structural frames, and are calibrated to those values. When the frequency of a subject building falls outside of that default range, common design
approaches provide for specified adjustments to the quasi-static design loads to
account for dynamic response.
Common seismic design (American Society of Civil Engineers 2005) involves
a very elaborate conversion of the dynamic loading environment into a quasistatic analysis problem. Building systems are characterized for stiffness and ductility, and site conditions are evaluated for seismic exposures and characteristics
of shaking. On the basis of extensive research into building performance and a
fair amount of cumulative experience evaluating the actual earthquake response
of designed structures, the complicated loadings—which are as much a function
of the building design as they are of the environment in which structures are
built—are idealized as a series of externally applied loads that are thought to
mimic the loading effects of an earthquake. Complicated though the approach is,
many buildings can be designed for earthquakes by engineers with little familiarity with dynamic behavior.
6
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
Our conventional approach to blast design is similar to that for seismic design,
in two important ways: (1) both loadings clearly are dynamic and, hence, solutions are energy-based, and (2) the way we detail structural elements determines
the effective loads for which structures must be designed (meaning, we limit the
strength we need to supply by allowing post-elastic behavior to dissipate energy).
However, blast loading, with its extremely fast rise time and usually short duration, is either dynamic or impulsive, depending on the nature of the explosive, its
distance from the subject structure, and the level of confinement that the structure
creates for the expanding hot gases (Mays and Smith 1995).
The impulsive form of the very fast load rise time and very short load duration normally associated with blast loading requires analytical approaches that
generally demand direct solution of energy balance equations (U.S. Department
of Defense 2008; Mays and Smith 1995).
1.4 STRUCTURE AS AN INFLUENCE ON BLAST LOADS
The pressure and duration of the impulse associated with a blast are influenced
by reflections of the shock front (U.S. Department of Defense 2008). Reflection
sources include the ground below the detonation point and building surfaces that
have sufficient mass or ductility to remain largely in place for the duration of the
impulse. When shock fronts are reflected, their pressures are magnified as a function of the proximity, robustness, and material characteristics of the impacted object (Bangash 1993). The more robust that object, the greater the reflected energy
because less energy is dissipated by the response (such as ground cratering) of
the surface. These variations often are neglected in conventional design.
For instance, facades normally are designed on the assumption that they are
perfect reflectors of the shock front. Designers following common procedures
are assuming that the facade components remain stationary for the duration of the
impinging shock front, causing peak pressures and impulses sufficient to reverse
the direction of the shock front. In practice, there can be some displacement of
the facade during the loading cycle. This displacement reduces the effectiveness
of the reflector, and correspondingly the impulse.
Analyses for interior explosions have additional complications, as designers
attempt to deal with the multiple reflections of the shock front within the structure, and pressures that develop from containment of expanding hot gases (Mays
and Smith 1995)–a phenomenon normally neglected for external explosions.
Further, the geometry of the confining volume and the location of the explosion within the volume can substantially affect the pressures on surfaces (U.S.
Department of Defense 2008). The science that describes the pressure history on
interior surfaces is complex, and not generally considered rigorously in common
blast-resistant design processes.
Providing blast venting through frangible components to mitigate the effects
of interior explosions is even more complex, since the release time for the venting
component is a key, but difficult to assess, factor in the determination of the
STRUCTURE AS AN INFLUENCE ON BLAST LOADS
7
magnitude of the pressure buildup. Approximations usually govern the analyses
(U.S. Department of Defense 2008).
Clearly, there is interplay between the performances of building facades and
frames. While in most cases the primary reason we enhance the performance
of a facade is to protect occupants, we gain protection for the structure as well.
Blast shock fronts that are not repelled by the facade will advance into a building,
inducing pressures on interior surfaces of the structure and threatening interior
columns, walls, and floor systems. Blast-related upward impulses on floor slabs
can reverse force distributions in these structural elements. In systems that are
not strong and ductile enough for these reversed forces, blast-induced deflections
can fracture structural elements that are required to resist gravity loads. Hence,
floor systems can fail after the direct effects of the blast pass and the slab falls
back downward under the influence of gravity.
Of course, by designing the facade to resist the effects of an explosion, the designer is forcing the structure to become a support for the blast loads. Depending
on the performance criteria, designers need to demonstrate that the framing system can support the applied loads, and that the structure as a whole will remain
standing with an acceptable level of damage.
Building enclosures normally are designed to resist blast effects by inelastic
flexural action, but it is possible to design facades to resist blast effects through
catenary action as well. In particular, blast retrofits sometimes include new
“catch systems” that are intended to reduce intrusion of blast pressures and creation of lethal missiles, by acting as a net inside the original exterior wall system.
In any case, the lateral displacement of the system often is large enough to
open gaps between wall panels or between panels and floor slabs. When this
happens, there is potential for leakage pressures to enter the building (U.S. Department of Defense 2002b), even when windows stay in place. This is particularly true in response to large, relatively distant explosions that have relatively
long-duration impulses.
Pressure fronts that leak past facades that are damaged but remain in place
normally are assumed to have insufficient energy to induce significant damage to
interior structural components. However, these leakage pressures can cause personal injuries and damage to architectural and mechanical systems if they are not
designed for resistance.
Add to the effects of leakage pressures the possibility that structural and architectural features on the inward-facing surfaces of facade components can
become missiles when the facade sustains damage as it deforms, and there remains substantial risk to occupants inside blast-resistant buildings even with
well-developed designs.
It is well established that breached fenestration leads to lethal missiles and
internal pressurization (American Society of Civil Engineers 1999). Common
design for blast resistance for malevolent attacks often is based on the premise
that a significant fraction of the fenestration in a building will fail (General Services Administration 2003). This is due in part to the variability of the properties
of glass, but also results from risk acceptance that employs the philosophy that
8
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
an explosion is unlikely and that full, “guaranteed” protection is prohibitively
difficult or expensive.
Hence, the effects of leakage pressures and missiles that are the product of
building materials fracturing in response to a blast often can be destructive to the
interiors of buildings, even when the facades of those buildings are designed to
resist the effects of an explosion. Except when the most restrictive approaches to
blast-resistant design are employed (e.g., with elastic response, so a building can
remain functional), parties with standing in the design process need to understand that substantial interior damage and occupant injuries are possible should
the design-base explosion occur.
1.5 STRUCTURAL RESPONSE
The shock front radiating from a detonation strikes a building component, it is
instantaneously reflected. This impact with a structure imparts momentum to exterior components of the building. The associated kinetic energy of the moving
components must be absorbed or dissipated in order for them to survive. Generally, this is achieved by converting the kinetic energy of the moving facade component to strain energy in resisting elements. Following the philosophy that blast
events are unusual loading cases that can be allowed to impart potentially unrepairable damage to structures, efficiency in design is achieved through post-yield
deformation of the resisting components, during which energy can be dissipated
through inelastic strain.
Of course, this means that the components that need evaluation often are deformed far beyond limits normally established for other loading types, and many
of the assumptions that form the basis for conventional design approaches might
not be valid. For instance, recognition of the extreme damage state normally
associated with dissipation of blast energy has led to debate about appropriate
values of the strength reduction factors ( factors) to be used for design.
In conventional design (American Concrete Institute 2005; American Institute of Steel Construction 2005), the nominal strengths of structural elements
are reduced by factors to account for uncertainty in the actual strength of
the elements, and for the consequences of failure. Their magnitudes for conventional design have been developed based on studies of structural responses
that are commensurate with service performance of buildings and, for seismic design, responses that are anticipated to be sufficiently limited and ductile to allow elements to retain most of their original load-carrying capacity.
Blast resistance, on the other hand, often takes structural elements far into
the inelastic range, to where residual strengths might be reduced from their
peaks, and alternative load-carrying mechanisms (e.g., catenary action) are engaged. Sometimes, designers anticipate complete failure of certain elements
if they are subjected to the design-base event. In this environment, it is not
at all clear that factors developed for conventional, nonblast design are
relevant.
NONSTRUCTURAL ELEMENTS
9
Common blast-resistant design often takes the values of the factors to be 1.0
(U.S. Department of Defense 2008). The bases for this approach range from the
uncertainty about what the actual values ought to be to the observation that loads
we assume for blast-resistant design are sufficiently uncertain that precision in
the values for is unjustified. It is further prudent to assume =1.0 when
performing “balanced design,” in which each structural element in a load path
is designed to resist the reactions associated with the preceding element loaded
to its full strength. Using =1.0 for determination of the full strength of the
elements in the load path tends to add conservatism to the loads required for the
design of the subsequent elements.
On the other side of the equation, designers often apply load factors equal to
1.0 to the blast effects (U.S. Department of Defense 2008). This follows from the
lack of a probabilistic base from which to determine the design threat, and the rationale that conservatism can be achieved by directly increasing the design threat.
In any event, the absence of complete agreement on how to address strength
reduction factors, and the valid observation that blast threats—particularly for
malevolent explosions—generally are difficult to quantify, reduce our confidence
in our ability to predict structural response with precision.
It is common in blast-resistant design to treat individual elements as singledegree-of-freedom nonlinear systems (U.S. Department of Defense 2008). Performance is judged by comparison of response to limiting ductility factors (i.e.,
the ratio of peak displacement to displacement at yield) or support rotations,
with the response calculated as though the structural element were subjected to
a pressure function while isolated from other structural influences. Of course,
much more sophisticated approaches are pursued for critical structures and complicated structural systems. However, research on structural response for very
high strain rates and very large deformations is limited, and results often are not
widely disseminated. In many respects, the sophisticated software now available
makes it possible to analyze with precision that exceeds our understanding of
structural response.
Hence, the simplified, single-degree-of-freedom approach forms the basis for
many designs. This approach usually is consistent with the precision with which
we model the blast environment and our knowledge of element behavior, but it
generally identifies the true level of damage only approximately. When considering elements as components of structural systems under the influence of blast, the
response of the individual elements can differ significantly from that determined
by analyses in isolation.
1.6 NONSTRUCTURAL ELEMENTS
Designers usually assume that the blast resistance of a structure is derived from
the elements that they design for this purpose. While this clearly is true in
large measure, in actual explosions, nonstructural elements—components disregarded in blast design—can act to reduce damage in a structure. It usually is
10
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
conservative, and therefore prudent, to ignore these components because the designer cannot be certain about the reliability, or even the long-term existence, of
building components that are not part of the structural design.
Nevertheless, elements with mass and ductility that stand between an explosion site and a target area can act to dissipate energy as they fail from the effects
of the blast. In fact, designers sometimes do rely on specific sacrificial elements
to reduce the blast effects on critical structural elements. The bases for this consideration are twofold: (1) through its failure, the sacrificial element dissipates
energy that would otherwise be imparted to the structural element, and (2) for the
brief time that the sacrificial element stays in place, it acts to reflect the shock
front, thereby reducing the impulse felt by the protected structural element. For
near-range conditions, when a bomb might otherwise be placed essentially in
contact with a key structural element, a sacrificial element such as an architectural column enclosure can enhance survivability simply by inhibiting close
placement of the explosive.
Of course, any shielding element that has inadequate strength, ductility, and
connection to remain attached to resisting elements is likely to become a missile. Some of the energy these elements absorb is dissipated through strain, but
the rest is retained as kinetic energy. The hazards created by these flying elements end only when that kinetic energy is brought to zero. Furthermore, care
is needed in the evaluation of the value of shielding elements that are not positioned closely to the structure under consideration, since shock fronts reform
beyond such elements, mitigating the protective value of the shield.
1.7 EFFECT OF MASS
The first influence of gravity comes to play when assessing the weights that
the designer assumes are present in the structure at the time of an explosion.
These weights, which are derived from the structure itself and its contents, act
concurrently with the explosion-induced loadings. As a result, they “consume”
some of the resisting capacity of the elements that are designed to resist the
explosion. In addition, for the most part, they remain on the structure after the
explosion and therefore must be supported by the damaged structure. The postblast distribution of these weights often will be uncertain.
On the beneficial side, mass often augments the blast resistance of structural
elements. Blast effects usually are impulsive, meaning that they impart velocity
to objects through development of momentum. With momentum being proportional to the product of mass and velocity (Eq. 1-1), and kinetic energy being
proportional to the product of mass and velocity squared, the larger the mass, the
smaller the velocity and, hence, the smaller the energy that must be dissipated
through strain (Eq. 1-2).
t1
(1.1)
I = F (t) dt = M V
t0
EFFECT OF MASS
where:
11
I = impulse
F(t) = time-varying force
t = time
M = mass
V = velocity
Ek =
1
1 I2
MV 2 =
2
2M
(1.2)
where: Ek = kinetic energy
Gravity also must be considered when elements or overall structures deform.
Vertical load-carrying elements often are designed to resist simultaneous vertical and lateral loads. Even when columns are not part of a structure’s lateral
load resisting system, it is common for them to be designed for an eccentricity
of the vertical load to account for inevitable moments that will develop in use.
Sometimes the magnitude of the eccentricity causing moment is assumed to be
on the order of 3% to 10% of the element’s cross section dimension (American
Concrete Institute 2005). Response to blast often deforms vertical structural elements far more than limits assumed for conventional design. The designer needs
to evaluate the ability to resist the resulting P- effects, both for individual elements and for the structure overall.
Structures as a whole generally are not pushed over by a common explosion.
The overall mass of a structure usually is large enough to keep the kinetic energy
imparted to the structure as a whole small enough that it can be absorbed by the
multiple elements that would need to fail before the building topples.
In many explosions that cause extensive destruction, the damage develops
in two phases: (1) the energy released by the explosion degrades or destroys
important structural elements, and (2) the damaged structure is unable to resist
gravity and collapses beyond the area of initial damage. In some of the most
devastating explosions, most of the structural damage has been caused by gravity
(Federal Emergency Management Agency 1996, Hinman and Hammond 1997).
Normally, individual elements fail, necessitating the activation of alternative
load paths within the structure to carry the gravity loads that remain after the
direct effects of the blast pass. Studies that assess these alternate load paths need
to consider the dynamic application of the redirected internal forces, as the sudden removal of load-carrying elements implies a change in potential energy, as
portions of a structure begin to drop. This change in potential energy necessarily imparts kinetic energy that must be converted to strain energy for the falling
mass to be brought to rest.
Hence, the evaluation of the full effect of a blast does not end with calculations
of blast damage to individual elements or limited structural systems. Designers
need to consider the ongoing effects in the damaged structure, under the influence
of gravity.
12
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.8 SYSTEMS APPROACH
In our efforts to enhance the blast resistance of a facility, we need to remain cognizant about how our designs affect the performance and viability of the facility
for nonblast events. As is always the case, there are competing goals and influences in the design of a facility, and those factors need to be balanced to achieve
the most satisfactory end product.
Consider the conflicts between the structural performance preferred for
seismic events and that preferred for explosions. One important goal in seismic
design is to force failures to occur in beams before columns, so that the
load-carrying capacity of columns is preserved even when the earthquake
induces damage. This is accomplished by detailing connections between beams
and columns so that plastic moments occur in the beams before the columns.
This is the “strong column, weak beam” approach.
Consultants designing for blast often provide for the possibility that a column
will be severely damaged by an explosion, in spite of our best efforts at prevention. When consultants assume that a column has lost its strength, they must develop alternative load paths to prevent a collapse from progressing from the initially damaged column through the structure. One form of alternative resistance
involves making beams strong and ductile enough to span over the area of damage, thereby redistributing the load on the damaged column to adjacent columns.
This requires strong beams which, if implemented without consideration of seismic response, can run counter to philosophies for robust seismic resistance.
Designers working to enhance blast resistance must also consider occupant
egress and the needs of emergency responders. Blast resistance invariably includes fenestration with blast-resistant glass. By definition, such glass is difficult
to break. Firefighters will need to use special tools and engage unusual tactics to
fight a fire in a building that is difficult to enter and vent, and that has features
that inhibit extraction of trapped occupants. Designers might need to compensate
for blast-resistance features or enhance fire resistance.
Distance is the single most important asset to a structural engineer designing
for blast resistance. The farther the explosion is from the structure, the lower
are the effects that the structure must resist. Further, there often is merit to the
construction of blast walls or line-of-sight barriers to add protection to a facility.
However, the need to create an impenetrable perimeter, and the temptation to
make it one that effectively hides the facility, can detract from the function of
the facility.
First, there is the dilemma caused by features that are intended to keep aggressors away from a building, but that also block lines of sight to the building in
the process. While such features add security, they also provide opportunities for
the aggressors to effectively hide from observers in and around the building. A
slowly developing assault may be more difficult to detect if the perimeter cannot
be monitored effectively.
Next, there is the potential impact on the quality of life for occupants of
buildings that have very robust defenses. Imposing perimeters and minimized
fenestration display the robustness and the fortresslike design intent. While
INFORMATION SENSITIVITY
13
this might be perceived as an asset for what it says to the aggressor, it also
communicates a sobering message to occupants and welcomed visitors. There
has to be a balance between the means to provide the necessary resistance
and the architectural and functional goals of the facility. Aesthetics need
consideration for most facilities.
Overall security design needs to properly balance the efforts applied to the
defense against a variety of threats. It is unsatisfactory to provide a very robust
design to resist blast if the real threat to a facility is through the mechanical system. Clients will be unhappy if security protocols address perceived threats (e.g.,
outside aggressors detonating bombs near a building), but fail to prevent real
threats (e.g., disgruntled employees intent on committing sabotage or violence
inside the facility). Any overall security evaluation needs to consider all perceived threats and provide guidance that will allow clients to determine where
best to apply their efforts to maximize their benefits. In many cases, a robust
resistance to an explosion threat will not be the best expenditure of funds.
Given a security design developed for the spectrum of potential threats to a
facility, owners sometimes face costs that exceed their means. When this occurs,
and for facilities that risk assessments show to be at relatively low risk, owners
must make decisions. Sometimes they instruct consultants to design to a particular cost, representing the amount that the owner can commit to the added security
to be provided to the facility. In these instances, consultants must identify priorities that address the most likely threats and provide the greatest protection for
the limited funds. When this happens, the consultants must explain to the owners
the limitations of the options so that the owners can make educated decisions.
1.9 INFORMATION SENSITIVITY
When blast-resistant designs are for the security and safety of a facility in response to a threat of a malevolent attack, information about the assumed size
and location of an explosion should be kept confidential. This information could
be useful to an aggressor because it can reveal a strategy to overwhelm the designed defenses.
The common practice of specifying the design loads on drawings should not
include a specific statement about the assumptions for blast loading when facility
security is at issue. Potentially public communications among members of the
design team and between the design team and the owner should avoid revelations
about the design-base explosion.
In most cases, the design assumptions for accidental explosions are not sensitive. Precautions about security-related confidentiality usually do not apply,
and customary processes for documenting the design bases may be followed.
In addition, there might be legal requirements or other circumstances that dictate the documentation of otherwise sensitive information. As always, designers
will need to comply with the law and to work with stakeholders in the design
of a facility to contain the unnecessary dissemination of information that could
potentially be misused.
14
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
1.10 SUMMARY
As consultants in the building design industry have been drawn into the matter of
blast-resistant design, many have been handicapped by lack of familiarity with
the blast environment, including not knowing how to determine loads for design,
or with proper approaches for structural design. Consultants often anticipate that
they will be able to provide effective designs by following approaches common
in building design when blast is not an issue. Unfortunately, consultants expecting to apply their familiar approaches usually are proceeding along an improper
path.
An explosion is a violent thermochemical event. It involves supersonic detonation of the explosive material, violently expanding hot gases, and radiation of a
shock front that has peak pressures that are orders of magnitude higher than those
that buildings normally experience under any other loadings. Designers hoping
to solve the blast problem by designing for a quasi-static pressure are likely to
be very conservative, at best, but more probably will simply be wrong.
Designers need to understand that the magnitudes of the pressures that an explosion imparts to a structure are highly dependent on the nature of the explosive
material, the shape and casing of the device, the size and range of the explosion, the angle of incidence between the advancing shock front and the impacted
surface, the presence of nearby surfaces that restrict the expansion of hot gases
or that reflect pressure fronts, and the robustness of the impacted surface itself.
Designers also need to understand that the durations of the pressures induced by
an external explosion generally are extremely short compared to the durations
of other loads and compared to natural periods of structures. Further, there is
interplay between blast pressure magnitudes and durations, which is a function
of distance from the detonation point, among other factors.
Designing for the very high peak pressures and short durations of blast loadings requires applications of principles of dynamic response. Accurate prediction
of the peak response of a building will require the designer to analyze dynamic
properties of the structure, and apply approaches that respect dynamic behavior.
Further, most cost-efficient designs rely on deformation far beyond elastic limits
to dissipate energy. Hence, many of the assumptions designers normally make
when designing for loads other than blast do not apply when designing for blast
resistance.
Consultants engaged in the design for blast resistance need to be qualified
by education, training, and experience to properly determine the effects of an
explosion on a structure. They must have specialized expertise in blast characterization, structural dynamics, nonlinear behavior, and numerical modeling of
structures. Blast resistance designers must be licensed design professionals who
are knowledgeable in the principles of structural dynamics and experienced with
their proper application in predicting the response of elements and systems to
the types of loadings that result from an explosion, or they must work under
the direct supervision of licensed professionals with appropriate training and
experience.
REFERENCES
15
The present practice for blast-resistant design employs many approximations
and, in many aspects, relies on incomplete understanding of the blast environment and structural behavior. While available approaches serve the public by
increasing the ability of our structures to resist the effects of explosions, these
conventional approaches generally are ill suited to provide a clear understanding
of the post-blast condition of the structure. Consultants providing blast-resistant
design need to understand the limitations of the tools they apply, and provide
clients with appropriate explanations of the assumptions, risks, and expectations
for the performance of blast-resistant structures. In many cases, those explanations need to make clear that the performance of the structure and the safety of
individuals inside the protected spaces are not guaranteed.
REFERENCES
American Concrete Institute. 2005. Building Code Provisions for Structural Concrete
and Commentary (ACI 318-05). Farmington Hills, MI: American Concrete Institute.
American Institute of Steel Construction. 2005. Specifications for Structural Steel Buildings. Chicago, IL: American Institute of Steel Construction, Inc.
American Society of Civil Engineers. 1999. Structural Design for Physical Security:
State of the Practice. Reston, VA: American Society of Civil Engineers.
. 2005. Minimum Design Loads for Buildings and Other Structures (ASCE/SEI
7-05). Reston, VA: American Society of Civil Engineers.
Bangash, M. Y. H. 1993. Impact and Explosion: Analysis and Design. Boca Raton, FL:
CRC Press, Inc.
Center for Chemical Process Safety. 1996. Guidelines for Evaluating Process Plant
Buildings for External Explosions and Fires. New York, NY: American Institute of
Chemical Engineers Center for Chemical Process Safety.
Federal Emergency Management Agency. 1996. The Oklahoma City Bombing: Improving Building Performance Through Multi-Hazard Mitigation (FEMA 227).
Washington, DC: Federal Emergency Management Agency.
. 2003. Reference Manual to Mitigate Potential Terrorist Attacks Against Buildings (FEMA 426). Washington, DC: Federal Emergency Management Agency, Department of Homeland Security.
General Services Administration. 2003. Facilities Standards for Public Buildings Service
(P100). Washington, DC: General Services Administration.
Hinman E. E. and D. J. Hammond. 1997. Lessons from the Oklahoma City Bombing:
Defensive Design Techniques. Reston, VA: American Society of Civil Engineers Press.
Mays G. C. and P. D. Smith. 1995. Blast Effects on Buildings. London: Thomas Telford
Publications.
National Research Council. 1995. Protecting Buildings from Bomb Damage: Transfer of Blast-Effects Mitigation Technologies from Military to Civilian Applications.
Washington, DC: National Research Council, National Academy Press.
Smith P. D. and J. G. Hetherington. 1994. Blast and Ballistic Loading of Structures.
Oxford: Butterworth Heinemann.
16
GENERAL CONSIDERATIONS FOR BLAST-RESISTANT DESIGN
U.S. Department of Defense. 2002a. DoD Minimum Antiterrorist Standards for Buildings
(UFC 4-010-01). Washington, DC: United States Department of Defense.
. 2002b. Design and Analysis of Hardened Structures to Conventional Weapons
Effects (DAHS-CWE) (UFC 3-340-01) (TM 5-855-1). Washington, DC: United States
Department of Defense.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002), Washington, DC: United States Department of Defense.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
2
Design Considerations
Robert Ducibella and James Cunningham
2.1 INTRODUCTION
On April 19, 1995, a Ryder truck containing about 5,000 pounds of ammonium nitrate fertilizer and nitromethane exploded in front of the Alfred P.
Murrah Federal Building in Oklahoma City, Oklahoma. The blast collapsed a
third of the building, gouged out a crater 30 feet wide by 8 feet deep, and destroyed or damaged 324 buildings in a 16-block radius. The bomb claimed 168
confirmed dead.
On September 11, 2001, terrorists crashed two commercial jets into the
World Trade Center twin towers. Another hijacked flight slammed into the
Pentagon, while a fourth was forced down by passengers and crashed in a field
near Shanksville, Pennsylvania. The total dead and missing numbered 2,992:
2,749 in New York City, 184 at the Pentagon, 40 in Pennsylvania, and the
19 hijackers.
More than any other acts of domestic or international terrorism, these two
attacks have forever changed the American building-sciences community’s relationship to the society it serves. Since the first skyscrapers appeared in the late
nineteenth century, Americans have come to expect commercial structures of
exceptional beauty and functionality. After Oklahoma City and September 11,
many ordinary citizens also assume that new buildings are designed to protect
people during explosions as well as other natural or man-made disasters. With
few exceptions, however, significant movement toward achieving the recent imperative of both commercial utility and explosive blast resistance is a work in
progress.
Design professionals have gained enormous experience with plans and models
that anticipate structural responses to gravity, wind, and seismic loads. Preventing or curtailing random acts of terrorism by identifying their probability of occurrence and potential consequences, however, falls outside the general practice
of structural engineering.
This chapter proposes an innovative and largely untapped approach to blastresistant-building security design, a new paradigm in which senior individuals
who have a breadth and depth of experience in the areas of site planning, civil,
17
18
DESIGN CONSIDERATIONS
structural, mechanical, electrical, fire protection, and vertical transportation engineering; architecture; code and egress consulting; site planning; and security
engineering collaborate on a total blast-resistant building security design. This
team approach should take into consideration security and antiterrorist strategies
that fundamentally affect site selection and building design. In considering the
design of blast-resistant buildings, the design professionals must partner to
become an effective security design team.
In more traditional building security design efforts, security professionals simply present the design team with the results of their risk assessment, and the
security planning component is assumed to be largely finished. For reasons expanded upon hereinafter, the authors urge instead that the design team collaborate closely with security professionals and security engineers throughout the
entire design process. In the iterative process of designing a blast-resistant facility, the architectural team should become a security design team, supported by
homeland defense, intelligence, security, law enforcement, and blast consultant
experts. Therefore, the term security design team as used in this chapter means
the multidisciplinary building sciences/security/explosives experts group of professionals described above.
It is a concept capable of implementation and proven to work in a wide range
of facility types and locations where occupants, assets, and business missions are
deemed worthy of protection.
2.2 A NEW PARADIGM FOR DESIGNING BLAST-RESISTANT
BUILDINGS, VENUES, AND SITES
The following paragraphs describe a structured framework for threat and vulnerability data gathering and for risk assessment. Security concepts such as design basis threat, consequence management, functional redundancy, building location, and critical-functions dispersal are explored. A brief checklist of security
design considerations is presented, and the reader is introduced to the design
principles and guidelines that are expanded upon in the handbook’s subsequent
chapters.
The suggested risk assessment model for blast protection has six parts:
1. A threat identification and rating, which is the security design team’s analysis of what terrorists and criminals can do to the target.
2. An asset value assessment, which represents how much the project’s people
and physical assets are worth and what the responsible parties will do (and
pay) to protect them.
3. A vulnerability assessment, which represents the attractiveness of the target, and areas of potential weakness and/or avenues of compromise.
4. A site-specific risk assessment, which is the product of these three studies.
A credible site-specific risk assessment is the single most critical factor
A NEW PARADIGM FOR DESIGNING BLAST-RESISTANT BUILDINGS
Threat
Identification
and Rating
(Step 1)
Vulnerability
Assessment
(Step 3)
Asset Value
Assessment
(Step 2)
19
Benefits Analysis
How mitigation options
change the vulnerability
and ultimately the risk.
Risk
Assessment
(Step 4)
Consider
Mitigation
Options
(Step 5)
Risk
Management
Decision
Cost Analysis
How mitigation options affect
the asset’s criticality
and ultimately the risk.
Figure 2.1 Risk Assessment Process (Adapted from FEMA 452)
of the security blast-design process; it is the basis and rationale for ensuring that the protective design strategies are incorporated in the multidisciplinary security/explosives/building sciences team approach that is stressed
throughout this chapter.
5. Mitigation options, based on the risk assessment as a foundation.
6. Risk management decisions, driven by the mitigation options and informed
by the available project resources.
Figure 2.1 shows the five steps leading to the risk management decisions, and
their interaction with each other.
To be useful in influencing the design, the risk assessment should be completed early, preferably during the preliminary planning and conceptual design
phase, but certainly no later than the completion of initial design documents.
Extensive unclassified information about explosive events is available from
the FBI; Department of State; Department of Defense (DoD); Department of
Homeland Security (DHS); Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF); the U.S. Armed Forces; and other U.S. agencies. The authors urge
the design team to obtain and use this information. Since this information is being constantly updated, we advocate that you use an Internet search engine such
as Google, Mozilla Foxfire, or Ask.com to conduct your own research, thereby
ensuring that your information is both current and relevant. For example, a quick
search on the terms “Department of Defense explosive events” results in over
2 million hits while “FBI explosive events” returns over 600,000 citations, and
“ATF explosive events” creates over 200,000 references.
20
DESIGN CONSIDERATIONS
There are numerous how-to guides that lay out systematic approaches
to secure facility planning, design, construction, and operation. These various methodologies—the Department of Defense CARVER process, Sandia’s
RAMPARTTM software tool, the National Institute of Standards and Technology (NIST)’s CET, and the Federal Emergency Management Agency (FEMA)
building security series are among scores of candidates1 —are all based on a
process of threat assessment, vulnerability assessment, and risk analysis.
The authors have adapted and somewhat modified FEMA’s Risk Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks Against Buildings
(FEMA 4522 ) and FEMA’s Reference Manual to Mitigate Potential Terrorist
Attacks Against Buildings (FEMA 426) as their risk assessment models to
present the risk management process. Admittedly, FEMA 452 is a somewhat arbitrary choice of exemplar, although it is becoming a de facto industry standard
among security professionals. There are many alternative methods—an Internet
search of the term “risk assessment for buildings” resulted in over 3.7 million
hits—but no matter which risk assessment version the design team selects, the
methodology should be:
r Specifically written for the building sciences community of architects, engineers, and professionals who design not only high-security government
facilities but private-sector structures as well.
r Intended to serve as a multi-hazard assessment tool of a building and its
site, but readily adaptable to focusing closely on the explosive threat.
r Organized with numerous checklists, tables, and memory aids that will assist the design team in determining threats, risks, vulnerabilities, and mitigation options.
r Proven effective through years of extensive use in real-world threat and vulnerability assessments.
1
The DoD’s CARVER risk assessment process is a mnemonic rather than a model. First developed for the U.S. Special Forces in Viet Nam to target enemy installations, CARVER
stands for Criticality, Accessibility, Recognizability, Vulnerability, Effect, and Recoverability.
CARVER has recently become hard to find on the Internet but is widely available through
the DoD, Homeland Security, ASIS International, or law enforcement agencies, among others. Sandia’s Risk Assessment Method—Property Analysis and Ranking Tool (RAMPART(tm))
is a software-based methodology for assessing the potential risks of terrorism, natural
disasters, and crime to buildings, particularly U.S. government facilities. Read more at
http://ipal.sandia.gov/ip details.php?ip=4420. NIST’s Cost-Effectiveness Tool for Capital Asset
Protection (CET) is a software-based risk assessment tool that building owners and managers can
use to protect assets against terrorist threats. CET is available without cost on the NIST Web site.
See http://www2.bfrl.nist.gov/software/CET/CET 4 0 UserManualNISTIR 7524.pdf.
2
FEMA, located within the Department of Homeland Security, is the U.S. government agency
tasked with disaster mitigation, preparedness response, and recovery planning. FEMA 452: Risk
Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks may be downloaded from
www.fema.gov/plan/prevent/rms/rmsp452.shtm. FEMA 452 is a companion reference to the Reference Manual to Mitigate Potential Attacks Against Buildings (FEMA 426) and the Building Design
for Homeland Security Training Course (FEMA E155). This document is also a useful companion
to the Primer for Design of Commercial Office Buildings to Mitigate Terrorist Attacks (FEMA 427).
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
21
r Available in unclassified form and preferably at little or no cost. FEMA
452, FEMA 426, and their companion manuals, for example, are available
for free from http://www.fema.gov/plan/prevent/rms/rmsp452.shtm.
Increasingly, the building sciences community is being challenged to incorporate high levels of security into the design of facilities, sites, and venues
that do not yet exist. While the risk assessment tools spreading throughout
the law enforcement, public safety, security, and building sciences communities can be extremely useful, almost all of them are geared to improving
the security of sites and structures that are already standing. Consequently,
it is essential that the risk assessment model that is selected has also been
used to assess future buildings and venues, and to create security plans for
virtual sites.
Traditionally, the building sciences community has prepared for natural disasters by following prescriptive building codes supported by well-established and
tested reference standards, regulations, inspections, and assessment techniques.
Many man-made hazards such as toxic industrial chemicals storage, and numerous societal goals such as life safety, have been similarly addressed. The building
regulation system, however, has only just begun to deal with the terrorist threat.
In the absence of high-quality regulatory guidance, the design team must fall
back on its own resources and expertise—always remembering that the nature of
the potential threat and the desired level of protection are equally important and
inseparable design considerations.
2.3 A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
Experts are quick to point out that terrorists almost invariably seek publicity
and sometimes monetary reward or political gain as well. It should be emphasized, though, that terrorism has a powerful appeal to many of the marginalized people of the world. Throughout human history, asymmetric warfare—in
this case, terrorism—has had an undeniable allure to some and the sympathy of
many more. Because it can be spectacularly effective, terrorism will be around
for the foreseeable future, hence the legitimate concern for designs to mitigate
its effects.
As George Santayana famously said, “Those who cannot remember the past
are condemned to repeat it.” What follows, therefore, is a brief survey of broad
trends in domestic and international terrorism, for the purpose of planning future
security measures in site selection and facility design by learning from history
(Santayana 1905, 284).
2.3.1 Terrorists’ Use of Explosives
Explosives continue to be the terrorist’s preferred weapon, since they are destructive, relatively easy to obtain or fabricate, and still comparatively easy to
move surreptitiously on the ground and by sea. Terrorists are also well aware
22
DESIGN CONSIDERATIONS
that explosives produce fear in the general population far beyond the geographical location of their intended target.
2.3.2 Vehicle-Borne Improvised Explosive Devices
The following case studies include just some of the vehicle-borne explosive devices that have been used in the past quarter-century.
April 18, 1983—The modern era of vehicle-borne improvised explosive devices dates from April 1983, when a massive truck bomb destroyed the
U.S. Embassy in Beirut. The blast killed 63 people, including 17 Americans. The attack was carried out by a suicide bomber driving a van, reportedly stolen from the embassy in June 1982, carrying 2,000 pounds of
explosives.
October 23, 1983—A Shiite suicide bomber crashed his truck into the lobby
of the U.S. Marine headquarters building at the Beirut airport. The explosion, the equivalent of 12,000 pounds of trinitrotoluene (TNT) and alleged
to be the largest truck bomb in history, leveled the four-story cinderblock
building, killing 241 servicemen and injuring 60. Minutes later a second
truck bomb killed 58 French paratroopers in their barracks in West Beirut.
September 20, 1984—A van loaded with an estimated 400 pounds of explosives swerved around several barricades and U.S. soldiers and penetrated
the relocated U.S. Embassy annex compound in East Beirut. The suicide
bomb exploded 30 feet from the building, killing 11, including 2 U.S. servicemen, and injuring 58.
February 26, 1993—World Trade Center’s Tower One’s underground parking garage was rocked by a powerful explosion. The blast killed 6 people
and injured at least 1,040. The 1,310-pound bomb was made of urea nitrate
pellets, nitroglycerin, sulfuric acid, aluminum azide, magnesium azide, and
bottled hydrogen—all ordinary, commercially available materials. The device, delivered in a yellow Ryder rental van, tore a crater 100 feet wide
through four sublevels of reinforced concrete.
April 19, 1995—Timothy McVeigh detonated an ammonium nitrate/fuel oil
bomb in front of the Alfred P. Murrah Federal Building in downtown
Oklahoma City, Oklahoma.
August 7, 1998—Two truck bombs exploded almost simultaneously at U.S.
embassies in two East African capitals, killing 213 people in Nairobi and
11 more in Dar es Salaam. Some 4,500 individuals, principally Kenyans
and Tanzanians, were injured. In May 2001, four men connected with
al-Qaeda, two of whom had received training at al-Qaeda camps inside
Afghanistan, were convicted of the killings and sentenced to life in prison.
A federal grand jury has indicted 22 men, including Osama bin Laden, in
connection with the attacks.
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
23
December 14, 1999—An apparent plot to bomb the Los Angeles airport was
disrupted when Ahmed Ressam, a Canadian of Algerian background, was
arrested at a United States–Canada vehicle border crossing in Washington
State. Ressam had nitroglycerin and four timing devices concealed in his
spare-tire well.
September 11, 2001—Al-Qaeda terrorists crashed two commercial jets into
the World Trade Center twin towers, and another hijacked flight slammed
into the Pentagon, while a fourth was forced down by passengers and
crashed in a field near Shanksville, Pennsylvania.
June 14, 2002—A powerful fertilizer bomb blew a gaping hole in a wall outside the heavily guarded U.S. Consulate in Karachi, Pakistan. The truck
bomb, driven by a suicide bomber, killed 12 and injured 51, all Pakistanis.
October 12, 2002—A bomb hidden in a backpack ripped through Paddy’s Bar
on the Indonesian island of Bali. The device was small and crude, but it
killed the backpack owner, likely a suicide operative. The bar’s occupants,
some of them injured, immediately ran into the street. Fifteen seconds later,
a second much more powerful bomb—estimated at slightly more than a ton
of ammonium nitrate—concealed in a white Mitsubishi van was detonated
by remote control in front of the Sari Club. This blast killed 202 and injured
another 209.
May 12, 2003—Attackers shot security guards and forced their way into three
housing compounds for foreigners in the Saudi capital of Riyadh. The terrorists then set off seven simultaneous car bombs, which killed 34 people,
including 8 Americans and 9 Saudi suicide attackers, and wounded almost
200 more. The facades of four- and five-story buildings were sheared off.
One explosion left a crater 20 feet across, while several cars and six or
seven single-family homes within 50 yards of the blast were destroyed.
August 5, 2003—A powerful car bomb rocked the JW Marriott hotel in central Jakarta, Indonesia, killing 12 people and injuring 150. Police believe
the suicide bomb, which severely damaged the American-run hotel, was
concealed inside a Toyota car parked outside the hotel lobby. The terrorist
group Jemaah Islamiyah is believed responsible.
August 19, 2003—Sergio Vieira de Mello, the U.N. special representative in
Iraq, and at least 16 others died in a suicide truck-bomb explosion that
ripped through the organization’s Baghdad headquarters. The bomb-laden
truck smashed through a wire fence and exploded beneath the windows of
de Mello’s office in the Canal Hotel in the late afternoon, destroying the
building and shattering glass a half mile away. The concrete truck was said
to have been chosen as a Trojan horse because of all the construction work
going on in the area.
November 8, 2003—Seventeen people, including 5 children, were killed and
more than 100 were wounded in an armed raid and suicide car-bomb attack
on a residential compound in Riyadh, the Saudi Arabian capital. The huge
24
DESIGN CONSIDERATIONS
blast from an explosives-laden vehicle came after gunmen tried to break
into the compound and exchanged fire with security guards. The bomb
could be heard across the capital, and Riyadh residents who lived miles
away said they felt their buildings shake.
March 11, 2004—The routine of a morning rush hour in Madrid, Spain, was
shattered when 10 simultaneous explosions occurred aboard four commuter trains. The attacks killed 191 and wounded more than 1,800. Thirteen improvised explosive devices concealed in backpacks were used, but
three failed to detonate and were recovered by the authorities.
October 8, 2004—The Hilton hotel in the Red Sea resort of Taba, Egypt, was
rocked by a large explosion that killed 29 people and injured at least 120
more. The powerful blast was caused by two separate suicide car bombs,
each containing about 200 kilograms of explosives, which crashed into the
lobby before detonating simultaneously.
March 2, 2006—A suicide car bomb killed 4 people and injured 50 others in a
parking lot adjacent to the Marriott hotel in Karachi, Pakistan. The hotel is
about 30 yards from the U.S. Consulate. Among the dead was David Foy,
an American diplomat believed to have been the target of the attack. Police
said they believed the bomber rammed a car packed with high explosives
into the diplomat’s vehicle. The force of the blast, the most powerful of its
kind in Karachi to date, lifted the victim’s vehicle into the air, hurled it over
a 7-foot wall, and left a crater 6 feet deep.
June 2 and 3, 2006—Canadian security and intelligence services arrested 12
men and 5 youths on charges of plotting to bomb targets in southern Ontario, including Parliament. The terrorist ring had attempted to procure 3
tons of ammonium nitrate.
June 29 and 30, 2007—A car driven by suspected al-Qaeda sympathizers
rammed into the main terminal of Glasgow International Airport, setting
off a large blaze when the explosive devices caught fire instead of detonating. The previous evening in London’s West End, an attempt to set off
two car bombs was unsuccessful when the mobile phone triggering devices
failed.
2.3.3 Person-Borne Improvised Explosive Devices
Person-borne improvised explosive devices (PBIEDs) are usually less powerful
than vehicle-borne improvised explosive devices, but they are potentially more
deadly because they are usually detonated very close to their intended victims.
In general, PBIEDs are targeted against civilian populations or at military personnel in civilian settings such as bus stops. The act usually has a political (not
military) purpose, since the intention is to kill, maim, and terrorize a civilian population, and the target areas have few or no security measures in place to prevent
bombings. For example, the so-called Second Intifada in Israel and the occupied
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
25
territories, which occurred between October 2000 and October 2006, saw 267
clearly identified suicide bomber attacks, most of which involved person-borne
explosives. More recent attacks in Iraq and Afghanistan continue this deadly
trend.
PBIEDs are generally delivered to the target in the following ways.
r The explosives are transported in a package or backpack, and the bomber
leaves the area before the explosion. The Madrid commuter attacks of 2004
and the Centennial Olympic Park bombing3 in Atlanta, Georgia, during the
1996 Summer Olympics illustrate this technique.
r The PBIEDs are worn by the attacker, who expects to die in the explosion.
(Suicide bombers, including female terrorists, are increasingly used to carry
out such attacks.) The Amman, Jordan, hotel bombings of November 2005
were suicide attacks, as were the numerous Intifada-inspired explosive-vest
attacks in Israel following the collapse of the Camp David II summit in
2000.
r The person-borne bomb may not always, in fact, be carried in the traditional
sense of the word, but may instead be placed in a wheelchair, a wheeled
computer case, or a small cart pulled by the aggressor. This delivery mechanism has the advantages of (1) depriving the security forces of the visual
cues presented by a person with a bulky object under his or her clothes and
(2) allowing the bomber to deliver more explosives. Consequently, its potential explosives design basis threat (DBT) value—a description of a specific
explosive agent, a quantitative value, and a mode of attack that has a degree
of likelihood of occurring—is significantly greater than that of body-worn
devices. Mitigating the threat of a PBIED is essential in facility designs,
as its successful delivery and detonation have the ability to invoke mass
casualty, disproportionate damage, and progressive collapse.
r The person who delivers the bomb may be unaware of his or her role in
the act of terrorism—for example, convicted Unabomber Ted Kaczynski’s
use of postal delivery workers. Express couriers, common freight carriers,
or other similar second- and third-party delivery agencies such as FedEx or
UPS might also be used. These bombs differ from the usual PBIED threat,
as they arrive cloaked in the legitimacy associated with an expected delivery mechanism and deny security personnel the opportunity to profile the
actual bomber. Using this delivery method, the explosive DBT value could
be equal to the carrier’s parcel weight and size management thresholds.
2.3.4 Locally Available Explosives
In assessing a threat, perhaps the most important question to answer is “Can the
enemy obtain weapons?” The recent wave of terrorist bombings and bomb plots
3
See http://en.wikipedia.org/wiki/Centennial Olympic Park bombing for details.
26
DESIGN CONSIDERATIONS
around the world clearly answers that question, but the apparent ease with which
powerful explosives are available is worth discussing.
Shortly after the Taliban retreated from Kabul in November 2001, a search
of an al-Qaeda safe house uncovered a bomb-making training center. Among
the documents seized were a detailed explosives instruction manual, a table of
explosive mixtures classified by strength, and a table comparing the power of
potential detonators such as acetone peroxide (Boettcher and Arnesen 2002).
Shoe bomber Richard Reed used triacetone triperoxide (TATP) in his unsuccessful December 2001 attempt to down an American Airlines flight en route
from Paris to Miami. In July 2005, however, the four London bus and subway bombers detonated homemade acetone peroxide explosives with deadly
effect.
The al-Qaeda safe house also yielded handwritten lists of formulas that included instructions on how to make RDX and a version of C-4, the powerful
military explosive used against the USS Cole in December 2000. The documents
suggest that al-Qaeda was developing its own variant of C-4, known commercially as Semtex, for use in a wide variety of bombs. Creating explosives from
the base chemicals, the training manuals noted, would avoid the security problems inherent in finding a supplier and in transporting the contraband into the
target country.
The ability to buy explosives components locally “gives more latitude, more
autonomy and possibly some degree of elusiveness,” according to Tony Villa,
an explosives expert who has worked extensively for the U.S. government
(Boettcher and Arnesen 2002). Al-Qaeda is “not just a bunch of guys climbing along some jungle gym and going through tunnels and shooting their guns
in the air,” notes David Albright, a nuclear weapons design and proliferation expert (Boettcher and Arnesen 2002). “These are people who are thinking through
problems in how to cause destruction.”
Al-Qaeda videotapes, discovered in Afghanistan after the Taliban’s defeat,
also show a level of bomb-making skill that could allow terrorists to arrive unarmed in the target city and to easily buy and mix the ingredients of a highexplosive device. One of the captured videotapes, a detailed how-to guide,
demonstrates the production from scratch of high-quality TNT and other bomb
components using easily obtained materials. The videotaped demonstration reinforces the written bomb manuals but also makes the training much more
effective.
“What we did see,” Albright said, “is that when we compared this information
on high explosives to the Internet that these [training manuals] are much more
polished. They really did work with these formulas, tested these formulas, and
developed a procedure of making these high explosives that led to effective high
explosives in a safe manner (Boettcher and Arnesen 2002).”
“The overarching point here is that they can pick any venue or target city,
arrive in that city and, based on the video tapes, construct a bomb using common
materials,” Villa explained (Robertson 2002).
A BRIEF HISTORY OF RECENT TERRORIST ATTACKS
27
2.3.5 Some Counterterrorism Considerations
Counterterrorism officials have drawn several lessons from these and many similar attacks:
1. Terrorists have always stuck to the operational maxim that what is simple
is best, so the experts believe that increased security measures have pushed
terrorists toward softer targets and easier-to-deliver but deadly vehicleborne bombs.
2. The knowledge needed to create bombs and biological/chemical weapons
is available in bookstores, in widely circulated training manuals, on
CD-ROM, and on the Internet. Armed with this information, even an amateur terrorist can buy enough commercially available chemicals to make
his or her contribution to global terrorism.
3. As Oklahoma City bomber Timothy McVeigh and the four July 2005
London transit jihadists have demonstrated, even novice terrorists can
purchase bomb-making components and mix the ingredients to create a
high-explosive device. For example, terrorists have combined acetone, hydrogen peroxide, and a strong acid such as hydrochloric or sulfuric acid
to make triacetone triperoxide (TATP), a powerful and highly unstable
explosive.
4. In the brief sample of bombing case studies above, bombers detonated their
weapons simultaneously or nearly simultaneously 29 times, including the
7 synchronized blasts in Riyadh and 10 in Madrid.
5. Explosive charges greater than 2,000 pounds were used in nine of the
bombings described above. Weapons weighing less than 1,000 pounds appear to have been used in eight cases, while the remaining seven explosive
devices were described in terms such as “powerful” and “massive” without identifying the actual amount of explosives. This brief sample suggests
that the continued use of large bombs carried in vans, SUVs, trucks, or
passenger vehicles reinforced for this purpose is likely.
6. Suicide targets were often at hotels, in restaurants, and on public
transportation—places where civilians had gathered.
7. Terrorists have planned phased explosions to target emergency responders,
civilians who have not yet been evacuated, and bystanders.
8. Aggressive military, intelligence, and law enforcement measures since
September 2001 have disrupted the terrorists’ communications and logistical support. These worldwide activities have degraded the radical jihadists’
ability to mount operations as sophisticated and coordinated as the hijackings of September 11 and the 1998 bombing of the embassies in East
Africa. As the message of jihad reaches more disaffected young men, however, the bombings may be smaller in scale but more frequent.
28
DESIGN CONSIDERATIONS
9. The terrorists will opt for continual low-level bombings, therefore, as part
of an ongoing campaign to inflict some “revenge”—less dramatic and less
traumatic than the twin towers—on the enemy and to prove their relevancy.
10. The terrorists sought out and exploited gaps where security was either ineffective or lacking—most notably in the case of the U.S. commercial aviation industry on September 11, 2001.
11. Today’s terrorists harbor no compunctions about killing or injuring innocent civilians.
2.4 COLLABORATING TO ANALYZE RISK
The first risk assessment probably occurred when a prehistoric couple heard
strange noises outside their cave; in the intervening millennia, thousands of risk
assessment formats have been developed. Today, in response to private-sector
concerns, the security community’s post-9/11 threat and vulnerability evaluations place far more emphasis on terrorism and criminal violence than they did
even a decade ago.
A comprehensive risk assessment for blast-protection design involves
close collaboration among city planners, architects, engineers, blast consultant subject matter experts, and law enforcement security professionals. In
this group effort, collaborating professionals assess and select the security
measures needed to detect, deter, prevent, defeat, mitigate, or recover from
terrorists’ bomb attacks.
This section introduces the analytical and decision-making steps that are
widely used today to identify the critical assets requiring protection, determine
the explosive threats, assess the vulnerabilities, and create mitigating measures.
It begins with a thorough but flexible five-step methodology for preparing and
executing a risk assessment.
2.4.1 Step 1—Threat Identification and Rating
The assessment process begins with a detailed understanding of the natural and
man-made threats that could realistically occur in the geographic area in question
and therefore pose a risk to the building or site being designed. To best prepare
the design team to anticipate the broadest range of contingencies, the threat identification and rating step calls for the following four tasks.
Task 1-1. Identifying the Threats. In blast-protection terms, a threat is a potential event culminating in an explosion that damages or destroys a protected
facility or its inhabitants.
Task 1-2. Collecting Information on Potential Attacks. Collecting information
for an explosives-oriented threat assessment requires expertise in weapons and
demolitions as well as in terrorist history, tactics, training, and targeting. An
informed estimate of the aggressors’ motivation and intentions against the future
COLLABORATING TO ANALYZE RISK
29
facility and its future occupants is also essential. The following list of questions
is a quick overview of the kinds of information the team needs to gather:
r What domestic or international terrorist groups are the potential aggressors,
and what are their capabilities?
r What has been expressed or can be inferred about their motivation and intentions against the clients?
r How skilled with demolitions are they? Are bomb-making materials available to them?
r Does the project have an economic, cultural, or symbolic significance that
terrorists could exploit to justify their acts and to gain publicity?
Many informative checklists and memory aids are available for collecting
these data, and several of the best have already been cited. The design team’s
security members should be well versed in at least one of these tools. They
should also have information sources among state and federal law enforcement
and homeland security agencies—if not by virtue of existing relationships, then
certainly through contacts and their own credibility. Neighboring building owners, operators, and tenants can also be a valuable data source.
Task 1-3. Determining the Design Basis Threat for Explosives. The Department of Defense defines the design basis threat (DBT) as the hazard against
which a building must be protected and upon which the protective system’s design is therefore based.4 In other words, the type and size of weapon the building
and site must be designed to withstand is the explosive DBT.
To calculate an explosive DBT, the team must first estimate the potential for an
attack by identifying the potential terrorists’ tactics and the tools, weapons, and
explosives they could employ. Terrorists have used explosives hidden in moving
and stationary vehicles and have left behind explosives that were hand-carried
to the target. Mail bombs and supply bombs—larger devices typically infiltrated
through shipping departments—and explosive devices carried by suicide terrorists are also proven techniques. For reference purposes, large-scale truck bombs
typically contain 10,000 pounds or more of TNT equivalent (all the following
bomb weights are in terms of TNT equivalent). Bombs in vehicles ranging from
small sedans to vans typically contain 500 to 4,000 pounds. Suicide bombers can
unobtrusively deliver belts ranging in size from less than 10 pounds up to about
40 pounds. Hand-carried explosives are typically on the order of 5 to 10 pounds
(Federal Emergency Management Agency 2005, 1–7).
Design documents do not normally specify the actual DBT explosive weights.
Instead, the weights are expressed in terms such as W-1 (for example, a 10-pound
bomb), W-2 (100 pounds), W-3 (1,000 pounds), and so forth. The actual values
may vary from project to project as determined appropriate by the design team.
4
This DoD definition of “threat” can be found at www.dtic.mil/doctrine/jel/doddict/data/d/01635.
html.
30
DESIGN CONSIDERATIONS
If design documents containing specific DBT explosive weights fell into terrorist
hands, they would become an extremely valuable aid to the enemy. Actual DBT
weights are revealed only on a need-to-know basis.
The willingness of suicide terrorists to turn themselves into bombs precludes
design professionals from adopting a design basis threat of “no explosives.” Consequently, when the design team is developing protective design strategies for a
site or building and its occupants, it must insist on a realistic cataloging of the
project-specific explosive DBTs. This information mining is essential if structural and architectural facility designs are to provide credible protection.
The structural engineers’ and architects’ protection strategies are based on
an estimation of the possible size and location of a detonation, coupled with a
realistic estimate of the facility’s ability to withstand an attack. Since the eventual design implementation concepts and components of a facility’s structure;
fac¸ade; evacuation, rescue, and recovery systems (ERR); and business continuity
measures are built upon this analysis, establishing quantifiable DBTs is a fundamental first step in providing the design team with a framework for designing
mitigation measures.
Task 1-4. Determining the Threat Rating. The team’s threat ratings will evaluate the probability of various blasts occurring and the likely consequences if they
do. The team should imbed probability and consequences in their consciousness;
they are two extremely important concepts that security design planning teams
far too often overlook.
Vehicle-borne bombs continue to be the terrorists’ favorite tactic, primarily
because trucks can carry potentially devastating explosive cargos. But terrorists
will exploit the weakest links in a building’s protective design; person-carried
bombs have shown that relatively small weapons can cause great damage to exposed and structurally vulnerable building interiors. This reality appears to argue for defending buildings against every conceivable explosive attack, but total
protection through design and engineering is impractical in terms of structural
hardening and benefit-to-cost ratio. A potential solution for the “protect everything” dilemma is discussed in later paragraphs, when the authors address the
complementary roles of building design, physical security, operational security,
and continual situational awareness.
Facility designs outside North America are, of course, also vulnerable to explosives threats. For work outside of North America, it is especially important to
develop a comprehensive risk assessment that is specific to the geographic area,
the project site, and the facility.
In addition to the person- and vehicle-borne IED delivery methods, the use
of standoff military ordnance, especially the use of shoulder-fired arms such
as the ubiquitous rocket-propelled grenade (RPG), remains a significant concern. Structural engineers and architects should give careful consideration to
these military-grade weapons as part of a competent protective design threatmitigating strategy. Also, major headaches can be avoided by resisting the tendency to overestimate the man-made threat. Ninja-like terrorists in attack helicopters have not appeared yet, and the two-millennium history of terrorism
COLLABORATING TO ANALYZE RISK
31
clearly demonstrates that the enemy realizes that the best route to success is very
often the easiest one.
The design team must provide project-specific information at the earliest
stages in which conceptual designs are adequately advanced. The architects’ and
designers’ timely participation in the initial risk assessment process will greatly
influence the credibility—or lack of credibility—of the project design team’s
analysis of potential vehicle and pedestrian access routes. In short, the determination of the degree to which the project site, physical plant, and personnel are at
risk from specific explosives must be informed by an interactive process of information exchange between the threat-assessment and design-team professionals.
The team, concluding its work on threat assessment, now produces a single
threat rating for the entire project. Threat ratings are commonly expressed in
a seven-level linguistic scale from very low probability to very high probability. “Very high” threats are considered credible and imminent, and “very low”
threats are deemed negligible. Between those two extremes, the spectrum includes threats of high, medium high, medium, medium low, and low probability.
The scale could be numerical instead, from 1 (very low) to 10 (very high). Some
threat-ranking formats result in both linguistic and numeric ratings. This veryhigh-to-very-low scheme is also used in ratings of asset value and vulnerability.
Figure 2.2 illustrates the use of worksheets to consolidate the design team’s
knowledge in a single, agreed-upon format. In this example, the worksheet displays the probability of various improvised and military explosives being used
in an attack and the anticipated damage as a consequence of the assault.
The authors encourage teams to use their own professional experience to question the raw scores produced by the threat-ranking scales. Applying a “Does this
make sense?” filter is wise, because rating threats is more art than science. Deferring to the client’s insistence that his facility is a likely terrorist target, for
example, puts the assessment process at risk of being driven toward the “very
high” end of the scale, resulting in overdesign. Of course, it is equally possible
to underrate the danger, and the consequences of underdesigning could be grave.
This threat identification and rating exercise creates a prioritized team assessment of the potential dangers, especially of a terrorist bombing, to the project.
2.4.2 Step 2—The Asset Value Assessment
The team will quickly need to learn as much about their building, its site, and the
surrounding area as they have discovered about the explosive threat. These assets
can be tangible, including facilities, equipment, inventory, and, of course, people.
Intangible assets include intellectual property, proprietary processes, reputation,
corporate information, and image. A concerted effort should be made to become
thoroughly acquainted with the client’s facility and its adjacent environment. The
team’s knowledge about those unique tangible and intangible assets that require
enhanced protection will be invaluable as the security design is developed.
In asset assessment, the first of four tasks is to identify the project’s layers of
defense.
32
DESIGN CONSIDERATIONS
Score
Collateral Damage at
Specified Distance
Site Population and
Capacity
Asset Accessibility
Asset Visibility and
Symbolic Value
Threat History for Buildings,
Tenants, and Environs
Knowledge and Expertise
Explosive Threats
Access to Explosives
Threat Determination Criteria
Improvised and Millitary Standoff Explosive Devices
1-lb. Mail Bomb
9
9
5-lb. Pipe Bomb
9
9
RPG or Similar Grenade/
Rocket Launcher
7
7
20-lb. Cargo Bomb
9
9
35-lb. Salchel Bomb/
Suicide Bomber
8
8
500-lb. Car Bomb
6
8
5,000-lb. Truck Bomb
4
8
20,000-lb. Truck Bomb
2
6
Natural Gas
2
8
NOTE: The threat numbers are national and are used for illustrative purposes only.
Figure 2.2 Primary Explosive Threats to Notional Urban Building (Adapted from
FEMA 426 and FEMA 452)
Task 2-1. Identifying the Layers of Defense. Layered facility protection is a
time-tested security design engineering approach. Defense-in-depth creates concentric rings of security, each increasingly difficult to penetrate, which provide
security and law enforcement officials with multiple opportunities for warning
and reaction. A layered defense complicates the terrorists’ operations, perhaps
to the point that security forces are able to interrupt the aggressors’ surveillance
or intercept them early in their attack. The obstacles presented by a layered defense may cause the terrorists to attack a less well-defended target, or could allow
building occupants time to relocate to predetermined defensive positions or find
shelter in designated safe havens during an assault.
COLLABORATING TO ANALYZE RISK
33
The first layer of defense in a typical urban setting is the surrounding
area—the sidewalks, curb lanes, streets, other buildings, and the neighborhood
itself. The structural composition of adjacent buildings can either enhance the
project’s security or increase its threat. Local law enforcement and traffic authorities are increasingly willing to support a security designer’s street and curb-lane
safety initiatives, particularly when realistic procedural measures precluding adjacent vehicle traffic lanes or parking are planned. Unobtrusive physical security devices—such as decorative anti-ram, crash-resistant street lights, benches,
planters, and trash receptacles—are found often now on sidewalks and plazas.
Embedded barricades, retractable bollards, and hydraulic barricades are also becoming more commonplace.
The exterior space between the line of vehicle or pedestrian interdiction and
the project structure is the second layer of defense. Surveillance cameras and
lighting must be part of any security design, but force protection engineered
vehicle interdiction devices and pedestrian approach restrictions can be useful
here too. Crime Prevention through Environmental Design (CPTED),5 a multidisciplinary approach to deterring criminal behavior, may also create a disincentive to terrorist surveillance and operations. ASIS International,6 the society
for professional security officers, has extensive libraries of security equipment,
concepts of operation, site civil and landscape counter-crime features, and operational procedures-related information to help in security design planning of this
interstitial area between the protected building and the site perimeter.
The third layer of defense is the building itself, and it is here that security
design measures can play a particularly effective role. Hardening structures, facades, and systems; providing a broad array of alarm and surveillance tools;
and carefully designing and locating utilities and building systems, incorporating equipment that supports evacuation and recovery, are examples of critical
elements in security design. Subsequent chapters of this handbook expand on
topics such as the role of materials-performance criteria, structural systems design, facade protection, and survivability of ERR systems.
Task 2-2. Identifying the Critical Assets. The second task of the asset value
assessment is to identify critical elements within each of the three defense
layers—in other words, which resources must withstand an attack if the health
and safety of the building’s occupants are to be ensured. FEMA and ASIS,
among others, publish extensive checklists that will greatly assist in pinpointing relevant issues.
Task 2-3. Identifying the Building Core Functions and Infrastructure. The security design team should next identify the building’s, site’s, or venue’s core
functions and infrastructure. The design team should use these data to inform
designs that isolate, distribute, or replicate the equipment and processes that will
5
A discussion of CPTED, including numerous references and external links, can be found at
http://en.wikipedia.org/wiki/Crime prevention through environmental design.
6
ASIS can be found at www.asisonline.org.
34
DESIGN CONSIDERATIONS
provide vital operations and services during and after an attack, or designs that
use hardening strategies to improve the operational survivability of these operations and services.
One of the most successful strategies for the protection of building core functions and infrastructure in blast-protection design is dispersal and redundancy. If
a bomb detonates adjacent to or within a facility, phones and telecommunications
equipment, security and evacuation systems, fire suppression and smoke control
systems, generators and uninterruptible power supplies, and other infrastructure
elements deemed critical to evacuation, rescue, and recovery must continue to
function. Here again, FEMA, ASIS, and others have developed extensive lists of
questions to help assessment teams identify core security and life-safety functions and their interdependencies.
Task 2-4. Determining the Asset Value Ratings. The final task in Step 2 is to
assign a value rating to each asset. The value rating is the security design team’s
judgment of the degree of debilitating impact that would be produced by the
incapacitation or destruction of that asset. In other words, the consequences of
damage or destruction are considered for each asset. Then a 1-to-10, very-highto-very-low value is assigned to each asset. “Very high” would mean that the loss
or damage of the building’s assets would have exceptionally grave consequences,
such as extensive loss of life, widespread severe injuries, or total loss of primary
services, core processes, and functions. On the other hand, a “very low” ranking
indicates that the loss or damage of the building’s assets would have negligible
consequences or impact. As was previously noted, there are five additional gradients between these two extremes (Federal Emergency Management Agency
2005, Table 2.5).
The asset value assessment lists the building’s core functions and critical infrastructure elements. Next, the team assigns each element an intrinsic value
based on the degree of debilitating impact that its incapacity or destruction would
create. The product of the value asset rating exercise is a table that lists the various types of attacks against each of the facility’s critical functions. The team
then populates the table with high-to-low values expressing the consequences of
loss or damage of each asset.
2.4.3 Step 3—The Vulnerability Assessment
Vulnerability is any weakness that aggressors can exploit to cause damage. To
constructively influence design considerations, the vulnerability assessment must
realistically appraise the building’s projected functions, systems, and site features; it must pinpoint structural weaknesses and identify absences of necessary redundancy so that corrective actions and mitigations can be incorporated.
Unlike threats, vulnerabilities are conditions and/or designs that people create
themselves and over which they can therefore exercise some control. Close coordination among the assessment team, the building design professionals, and
ownership is critical; it is essential to prevent a great deal of incompatible expectations and design deliverables down the road.
COLLABORATING TO ANALYZE RISK
35
There are four tasks in assessing vulnerability:
Task 3-1. Organizing Resources to Prepare the Vulnerability Assessment.
Task 3-1 has two subparts:
(a) Selecting the Assessment Team. Risk assessment is a complex undertaking with high stakes in terms of consequences and costs. Proper team selection requires some serious up-front time and attention, if only because,
in unskilled hands, risk assessment tools can be blunt instruments. The
best survey-tool authors stress that “teams created to assess a particular
building should be senior individuals who have a breadth and depth of
experience in the areas of civil, electrical, and mechanical engineering;
architecture; site planning and security engineering; and how security and
antiterrorist considerations affect site and building design” (Federal Emergency Management Agency 2005, ii).
(b) Determining the Level of the Assessment. The security community generally recognizes three increasingly detailed levels, or tiers, of risk assessment (explained below), so at this point a decision is required: how
detailed must the evaluation be?
FEMA 452 guidelines for vulnerability assessment denote three assessment
levels:
Tier 1—A study that identifies the most important vulnerabilities and their
mitigation options—a “70 percent” assessment. A Tier 1 evaluation is typically conducted in approximately 2 days by one or two experienced assessment professionals working with the building ownership and key staff.
It involves a quick look at the site perimeter, building, core functions, infrastructure, and any related drawings and specifications. A Tier 1 assessment is likely sufficient for the majority of commercial buildings and other
noncritical facilities and infrastructure.
Tier 2—A full on-site evaluation by assessment specialists that provides a
robust evaluation of proposed and/or existing systems’ interdependencies,
vulnerabilities, and mitigation options—a “90 percent” assessment solution. A Tier 2 assessment typically requires three to five assessment specialists. It can be completed in about 5 days and requires significant key
building staff participation; access to all existing site buildings and areas,
systems, and infrastructure; and/or an in-depth review of proposed building design documents, drawings, and specifications. A Tier 2 assessment
should be sufficient for most high-risk buildings such as iconic commercial buildings, government facilities, cultural and educational institutions,
hospitals, transportation infrastructure, and other high-value targets.
Tier 3—A detailed evaluation of the facility and site. A Tier 3 analysis uses
blast and weapons-of-mass-destruction (WMD) modeling to assess the site
and building’s response, survivability, and recovery parameters. It involves
36
DESIGN CONSIDERATIONS
engineering and scientific experts, requires detailed design information,
and provides the maximum data for evaluating and developing mitigation options. Tier 3 modeling and analysis can often take several weeks
or months for an existing facility and is a process involved at the Conceptual, Schematic, and Design Development phases of new facilities. It is
typically performed only for high-value and critical structures. From 6 to
12 subject matter experts may be required, based on project complexity.
When the building and site are in the design phase and do not yet exist, the risk assessment data-gathering process described above must be
modified considerably. A Tier 2 risk assessment of a future building would
concentrate, for example, on conceptual design documents, drawings, and
plans in an iterative process of discussion, impact assessment, and design
or redesign. This situation is actually preferable, because it gives the design
professionals an opportunity to build security into the structure’s skeleton.
The alternative—retrofitting protection into inappropriate or inadequate
spaces—seldom offers totally satisfactory solutions.
The choice of the appropriate tier level depends on the building and
site’s location, its construction characteristics, its proposed use and profile,
and its owners’ or occupants’ concerns about terrorism. The choice of tier
level is also heavily influenced by the design team’s initial threat and vulnerability appraisals; by local, state, and federal antiterrorism and public
safety regulations and guidelines; and (inevitably) by benefit-versus-cost
considerations.
Task 3-2. Evaluating the Site and Building. The authors have described how
producing a robust building security design requires the team to identify
potential threats and vulnerabilities, define the resources and people most
at risk, and describe the consequences of potential losses. To bring order
and predictability out of such complexity, the team’s next task is to prepare
a survey schedule and tentative agendas. The team needs, for example,
to meet with stakeholders such as site or building owners, security and
engineering personnel, IT directors, and emergency managers. The team
will also need to inspect the area and facilities and review key information
such as security plans and emergency procedures, if these documents exist.
The key to timely execution and success is planning.
An important aid to planning is threat mapping, the clear and graphic
representation of the site-wide explosive DBTs at vulnerable points, such
as lobbies, mailrooms, and loading docks. Threat maps, which are based
on architectural concept drawings, should encompass the site from its outermost perimeter to its most sensitive internal areas. This mapping can be
done on existing blueprints or other site drawings. Each vulnerable area is
labeled with a “W” number, according to the key that has already been created for designating explosive DBTs (W-1, W-2, W-3, etc.). Threat maps
illustrate for design professionals how concentric layers of competently applied security measures can reduce, or in some cases largely eliminate, the
facility’s explosive threats.
COLLABORATING TO ANALYZE RISK
37
Task 3-3. Preparing a Vulnerability Portfolio. It just makes good sense that
the large volume of data that has been collected as part of the threat, risk,
and vulnerability assessment in the previous steps should next be gathered
into a well-organized security design document.
One of the most useful sets of assessment aids in this team portfolio
can be the FEMA Building Vulnerability Assessment Checklists found at
Appendix A of the previously mentioned FEMA publication 452, Risk Assessment: A How-To Guide to Mitigate Potential Terrorist Attacks. This
series of tables prompts the team to adopt a consistent approach to evaluating security designs at their various stages of development. The checklist
can be used as a screening tool for preliminary design vulnerability assessments and supports the preparation of each step in FEMA’s How-To Guide.
As the FEMA Guide explains:
The checklist is organized into 13 sections: 1) site, 2) architectural, 3) structural systems, 4) building envelope, 5) utility systems, 6) mechanical systems, 7) plumbing and gas systems, 8) electrical systems, 9) fire alarm
systems, 10) communications and information technology, 11) equipment
operations and maintenance, 12) security systems, and 13) the security master plan. To conduct a vulnerability assessment or preliminary design, each
section of the checklist should be assigned to an engineer, architect, or subject matter expert who is knowledgeable and qualified to perform an assessment of the assigned area. Each assessor should consider the questions and
guidance provided to help identify vulnerabilities and document results in
the observations column. If assessing an existing building, vulnerabilities
can also be documented with photographs, if possible. The results of the 13
assessments should be integrated into a master vulnerability assessment and
provide a basis for determining vulnerability rating during the assessment
process. (Federal Emergency Management Agency 2005, Appendix A)
Task 3-4. Determining the Vulnerability Rating. The weaknesses the security
design team has identified, which could be exploited during an explosives
attack, are reflected in the vulnerability rating. As an example, the lack of
critical-systems redundancy raises the vulnerability rating, of course, as do
single points of failure among essential security and life-safety systems or
structural systems incapable of withstanding the project explosive DBTs
or an unforeseen event that could compromise a key primary structural
element.
Like the previous threat and asset value ratings, vulnerability scores are commonly expressed in a 1-to-10 numerical scale or a very-low-to-very-high linguistic scale. A very high vulnerability rating is delivered from identifying one
or more major weaknesses that make the facility extremely susceptible to an
aggressor or hazard. A very low rating describes a building that integrates excellent physical security and comprehensive redundancies—in short, a building that
38
DESIGN CONSIDERATIONS
would promise a high degree of physical and intellectual asset protection and be
operational immediately after an attack.
To summarize where the discussion is now in the risk assessment process:
In light of the identified potential threats (Step 1) and the team’s categorization
of the structure’s critical assets (Step 2), the vulnerability assessment (Step 3)
determines the opportunities to place at risk the project assets, and therefore sets
the stage for (Step 4) the risk assessment, wherein the team and the client discuss
and agree to the extent of acceptable risk.
The vulnerability assessment is a comprehensive listing of all weaknesses an
aggressor could exploit to damage an asset. The product of Step 3 is another
table, this one listing the building functions, each one with its high-to-low vulnerability rating, together with the associated vulnerabilities that the team has
uncovered.
2.4.4 Step 4—The Risk Assessment
This is the step in which the team pulls together all the information it has compiled, all the tables it has completed, all the judgments it has made, and each of
the threat, asset, and vulnerability ratings it has carefully assigned. This step is
where the interdisciplinary teamwork pays off; the risk assessment, backed up
by its demonstrable research, provides a credible foundation for security design
decisions.
Task 4-1. Preparing the Risk Assessment Matrices. Focusing attention on explosive event protection naturally narrows the range of security-design considerations. If, however, the team wants to undertake an all-risk, all-hazard review, including chemical/biological/radiological attacks as well as cyberassaults, it can
refer to FEMA 452 for a range of helpful checklists. Whether the analysis is
broad or narrow, however, the way to estimate potential losses is to prepare a series of matrices, tailored to the individual task at hand and based on the analysis
conducted during Steps 1, 2, and 3.
To estimate risk, the team must have analyzed a number of factors. First, it
will have identified and ranked the threats that could harm the project. Next,
it will have determined the value of the client’s assets, including people who
require protection. Finally, by using the results of the threat scores, asset value
estimations, and vulnerability ratings, it will determine a vulnerability and asset
rating system that identifies weaknesses an aggressor could exploit.
Task 4-2. Determining the Risk Rating. The design team next analyzes the
threat ranking, asset value, and vulnerability rating developed in the previous
steps and assigns a risk level for each critical asset. A specific risk such as a car, truck-, or person-borne bomb can be expressed as the sum of the asset value
times the threat rating times the vulnerability rating, or R = AV × TR × VR.
This exercise also results in linguistic or numeric values, or both.
Task 4-3. Prioritizing Vulnerabilities with FEMA’s Building Vulnerability Assessment Checklist. With a risk level now assigned to each critical asset, the team
is ready to prioritize, or rank, the vulnerabilities it has identified. Prioritization is
COLLABORATING TO ANALYZE RISK
39
based on evaluating which vulnerabilities the aggressors will most likely exploit
and which of them pose the greatest danger to lives, structures, and operations.
The highest ranking is given, obviously, to the greatest vulnerabilities with the
greatest consequences. Doing so makes it possible, in the next step, to determine
the most effective mitigation measures.
2.4.5 Step 5—Considering Mitigation Options
After completing the risk assessment, stakeholders are frequently left with issues
requiring mitigation, but they find their options limited by space constraints, esthetics, regulatory requirements, or costs. Decisions about allocating resources
therefore must focus on the most practical mitigation options.
These are the fundamental mitigation questions:
1. What kind of action should be taken in response to the danger? There are
four ways of responding to a threat:
r Acceptance of a threat is a rational alternative that is often chosen when
the threat has a low probability, low consequence, or both.
r Prevention is the alteration of the target or its circumstances to diminish
the risk that the terrorist or criminal will succeed.
r Interdiction is any confrontation with, or influence exerted on, an attacker to eliminate or limit its movement toward causing harm.
r Mitigation is preparation so that, in the event of an explosion, its consequences are reduced.
2. Does the response create new risks to the asset or others? The final step
in analyzing the security program’s efficacy is to be aware of new risks
created by the prevention, mitigation, or interdiction of the threats under
consideration.
Four tasks are involved in identifying mitigation alternatives:
Task 5-1. Identify Preliminary Mitigation Options. This exercise identifies the
design team’s available design options together with any relevant physical,
technical, cyber, and procedural security alternatives. The goal is to create
the strongest possible security design.
Task 5-2. Review Mitigation Options. Since every project has budget constraints as well as operations and maintenance limitations, it is seldom
possible to totally eliminate a risk. Now is the time, therefore, to weigh
mitigation options for their practicality and to match the most feasible mitigation opportunities with the most critical risks.
Task 5-3. Estimate the Cost. Some costs of protecting buildings are fixed,
while others are variable. Funds for deploying protective equipment and
creating a security zone-of-control are generally fixed; for example, the
40
DESIGN CONSIDERATIONS
costs of anti-ram barricades to keep a vehicle away from the facility are
comparable whether the same vehicle contains 50 pounds of TNT or 500.
The blast-protection costs of facility construction, on the other hand, will
vary according to the estimated threat level. Since the bomb size is outside
building planners’ control, the tailoring of blast protection to the building
structure, for instance, is a function of the maximum credible explosive
charge plus its estimated proximity when the bomb detonates.
Since cost is almost always a central issue, there are three mitigation approaches that can be used singly or in combination: (1) modify the design threat,
(2) reduce the level of protection, or (3) accept the risk. Life-cycle costs need
to be considered as well. For example, the recurring costs of maintaining and
upgrading security and life-safety equipment can derail even the best protective
programs.
Security design experience has repeatedly shown that attention paid to defending against man-made hazards during the preliminary planning and conceptual design phases significantly reduces the life-cycle costs of protection and
increases the inherent security level provided to the building occupants.
In summary, the project security professionals’ explosive threat assessment
should figure prominently in the design of the facility facade, structural system,
ERR systems, and other critical protective features. The design team should be
directly involved in the risk assessment steps that identify which building design
features and components can affect the type, size, and potential delivery method
of an explosive device.
Task 5-4. Reviewing Mitigation, Cost, and the Layers of Defense. FEMA 452
includes a final series of tables containing examples of mitigation options
and alternatives for each of the facility’s security features. These security
measures are organized by layers of defense and arrayed by degree of protection, cost, and effort. See Figure 2.3 below for FEMA 452 (January
2005) Worksheet 5-1: Preliminary Mitigation Options.
When the team has completed the risk assessment process, it will have identified and evaluated various mitigation options it can use as guidance when developing the building’s protective design features and components.
2.4.6 The Continuing Role of Risk Management
The goal of stronger, safer structures begins with building security design and
continues as security risk management throughout the building’s life cycle. To
support these initial and long-term goals, FEMA 452 asserts that the goal of the
assessment process is to achieve the level of protection sought through implementation of mitigation measures in the building design. These measures may
reduce risk by deterring, detecting, denying, or devaluing the potential threat
element prior to or during execution of an enemy attack. The Department of
Observation 1
Observation 2
Observation 3
Observation 4
Observation 5
Observation 6
Observation 7
CBR Protection and Control Measures
Blast Protection and Control Measures
CBR Repair and Strengthening of Existing
Structures
Blast Repair and Strengthering of Existing
Structures
CBR Regulatory Measures
Prioritized
Observations
Blast Regulatory Measures
COLLABORATING TO ANALYZE RISK
41
Worksheet 5-1 will help to
identify your preliminary
mitigation options.
After you have priortized your
observations/vulnerabilities
(Task 4.3), proceed to rank
them for impact during blast
and CBR events. Using the
first part of the Worksheet
(Prioritized Observations)
indicate if these observations
merit a regulatory,
rehabilitative, and/or
protective measure and if they
are directed at blast or CBR,
Using the second (Preliminary
Mitigation Options for Blast)
and third (Preliminary
Mitigation Options for CBR)
parts of the Worksheet,
determine mitigation options
that address the main
concerns included in your
observations and provided
parameters.
Preliminary Mitigation Options for Blast
Mitigation 1
Mitigation 2
Mitigation 3
Mitigation 4
Preliminary Mitigation Options for CBR
Mitigation 5
Mitigation 6
Mitigation 7
Mitigation 8
Figure 2.3 Worksheet 5-1: Preliminary Mitigation Options (From FEMA 452, Jan.
2005)
Homeland Security endorses the use of the following methodology (available in
FEMA 452).to achieve this purpose.
Deter: The process of making the target inaccessible or difficult to defeat with the
weapon or tactic selected.
Detect: The process of using intelligence sharing and security services response to
monitor and identify the threat before it penetrates the site perimeter or building
access point.
42
DESIGN CONSIDERATIONS
Deny: The process of minimizing or delaying the degree of site or building infrastructure damage or loss of life or protecting assets by designing or using infrastructure and equipment designed to withstand blast and chemical, biological, or
radiological effects.
Devalue: The process of making the site or building of little or no value or consequence, from the terrorists’ perspective, such that an attack on the facility would
not yield their desired result (FEMA 452, 2005).
2.5 CONSEQUENCE MANAGEMENT
This section provides insight into one of the primary tools used in assisting design professionals and their clients in the development of decisions involving
risk management. This section explores the value and process of evaluating consequences associated with explosive events and the major mitigating strategy
categories used to reduce those consequences.
Consequential effects are, by their very nature, subjective evaluations and categorizations. The perception of what is and what is not valuable and/or essential
is very much determined by the valuation process utilized by the institution or
organization that is responsible for its codification. Consequently, institutional
business mission statements, private organizational charters, government agency
programs, stockholder expectations, legal proscriptions, and numerous other
guidelines and/or mandates often identify different levels of expectation for postevent facility performance. These guiding documents and/or unwritten notions
were generally not created for the purpose of assisting design professionals by
clearly defining ownership’s requirements for preserving institutional businessmission operations or the protection of people on premises. The absence of wellcodified protection expectations leaves the design team with less than adequate
information about the extent of consequential effects that ownership will find
tolerable.
Design professionals must clearly understand the goals and expectations
of the facility, its occupancy and the activities that will take place within
the structure, site, and, in some cases, the campus or regional landscape
or cityscape fabric they are creating for (and with) their client. This is an
essential piece of early project information client data mining, which is often overlooked. Insufficient efforts and pursuit of a clear definition of client
desires regarding consequence management are one of the most significant
design process shortfalls. Mismanagement of this process will likely predestine the engineering efforts to fail to meet client needs or expectations.
To address this omission in the programming design phases, one of the first
steps should be—during the earliest phases of project design programming and
conceptualization—the creation of a brief security design mission statement.
This document should clearly and concisely state the client’s expectations for
the facility’s post-explosive event performance. These expectations are routinely
overstated by clients. This is generally a result of their lack of sophistication or
CONSEQUENCE MANAGEMENT
43
knowledge of the engineering complexities, architectural program compromises,
post-construction institutional operational constraints, impact on project financials, aesthetic effects, and other factors better understood by the structural engineer and the balance of design team professionals. Consequently, this mission
statement must be a consensus-developed document. It is reasonable to expect
that it may take several iterations of authorship, editing, review, and subsequent
approval.
Once agreed to, the consequence management statement should then become a part of the project’s key design assumptions and design program
objectives. It should be used as a means to guide design decisions, and to
provide the basis for decisions regarding program conformance, as designs
mature in the Schematic, Design Development, and Construction Document
phases.
Once a clear concept of acceptable consequence management in terms of facility performance is understood and codified, there is a sound basis for developing the facility designs. This is essential to the process of achieving an acceptable
balance between post-explosive event damage to the facility’s physical and intellectual assets and the expectations of the client and society regarding life safety
and institutional business-mission continuity.
No facility can be expected to provide a post-explosive event damage performance profile that ensures no injury, no loss of life, and no disruption to
facility operations. The challenge of consequence management is for the
design professional and client team to establish reasonable goals and objectives. These need to be tempered by rational opportunities for success
given the facility’s design program, client occupancy and activities, project
location and budget, law enforcement and security resources, and other economic and geopolitical factors, which are unique and specific to each design
commission.
Design professionals and their clients have opportunities to examine and implement strategies that can participate in achieving the level of consequence management identified in the project’s key design assumptions and design program
objectives, as stated above. These holistic design strategies usually include:
(a) Strategic site selection and facility location to reduce vulnerability to
threats that are on-site, near-site, or regional.
(b) Within a specific building, the provision of equipment, personnel, and resource redundancies, so that the loss of a single critical system, facility
management operational component, or structural element or system does
not create a single point of unacceptable or catastrophic failure.
(c) Within a specific building, distribution of facility physical plant; intellectual and nonintellectual assets; operational programs; and evacuation,
rescue, and recovery systems so that an explosive event in one location does not jeopardize the entire building, all of its occupants, and the
facility-wide performance of ERR systems. (Note, however, that the use of
44
DESIGN CONSIDERATIONS
diverse redundancy as a protective strategy should be examined during the
design process in conjunction with the protective strategy of co-location of
critical systems and/or resources housed within a well-protected environment. The intent is to foster survivability, thereby reducing consequential
effects.)
2.5.1 Consequence Evaluation
It is universally understood by the architectural and engineering design community that explosive events can be responsible for significant damage to building
structures, facades, and critical building systems. However, there is less awareness amongst the design profession during the initial planning phases of other
extremely important and sometimes ethereal values such as an institution’s public image and sense of solidarity. Without this awareness there is naturally less
acceptance of design responsibility for the protection of these other, less tangible
but extremely important client assets.
Additionally, other more essential issues associated with evacuation, rescue,
and recovery operations generally are assumed by the design profession to be
owner and operator responsibilities developed by those entities in response to
the building’s design. This approach is archaic.
Design professionals must address building designs so that the explosive
event consequence management response is achieved through an evaluation
of how evacuation, rescue, and recovery operations will occur after the explosive event. In this way, designs are informed to protect evacuation, rescue, and recovery building design features such as stairways, exit passageways, elevators, and building systems such as emergency power, emergency
lighting, voice evacuation, fire detection, fire suppression, and smoke management. This consequence evaluation approach results in a building design
that provides inherent tactical support and definitive criteria for an evacuation, rescue, and recovery concept of operations. This will always provide
a superior product to an ERR strategy retrofitted to a design created with
less inherent vulnerability.
The above design shortfalls can generally be attributed to the design team’s
focus on the intimate knowledge of and extensive involvement in developing
all of the intricate details associated with each component of the building’s architectural, structural, and support system infrastructure designs. Furthermore,
the design community traditionally focuses on contract boundary limits as the
extent of their responsibility for the ultimate scope of their work. The range
of effects from explosions rarely aligns with contract boundary limits of
work. This fact must be taken into consideration as part of consequence
evaluations.
Facility occupancy after construction completion, with its complement of staff
population assignments, support furnishings and equipment, and operational protocols usually is established with the greatest attention to detail after the building
is commissioned. Consequently, these and other aspects of day-to-day facility
CONSEQUENCE MANAGEMENT
45
management are logistically more distant entities and often are understood in
less detail by the design professional team. Design professionals should focus
additional resources on more completely understanding post-construction
occupancy and fit-out subtleties, in order for explosive event consequence
evaluations to be more thoroughly responsive.
Also, traditional contracts, governing the performance of the design team’s
scope of work, emphasize the most significant levels of effort up to and inclusive
of building commissioning. They fall short of emphasizing the role of the design
professionals in fully understanding a broader range of construction completion
and subsequent facility occupancy concerns. These include all of the various intellectual and operational entities, exceptional conditions and crisis management,
and facility restoration issues, which will be placed at risk and/or hindered by an
explosive event once the structure is completed. Understanding how the facility will operate under extraordinary conditions is essential to developing
protective designs that address the consequential effects on occupants and
operations, as well as on the physical plant.
The structural engineering, architectural design, and balance of the design team professionals should ensure, at the commencement of the design
process, that this disparate intellectual appropriation of design awareness is
addressed. This should be accomplished through a process that focuses on
the protection and post-explosive event performance of the structure itself
in a manner that is carefully balanced against all of the other consequential explosive event effects, including day-to-day facility operations and the
exceptional conditions associated with ERR.
A wide range of consequences must be evaluated. Understanding these consequences is critical to establishing the balance between (a) protective designs
assigned to the facility’s site perimeter protection program, facade, structure,
evacuation, and rescue and recovery systems and (b) systems associated with
life safety and business continuity. Mitigating each of the consequences that
may occur if these areas of facility vulnerability are not adequately addressed
will require a portion of the project’s budget. The project profile for security
investments will in large part be determined by the consequence evaluation
process. This process can be logically approached and informed by an evaluation of the following areas of explosive event influence:
Site and Local Risk Management Context Although design professionals
are traditionally charged with protective designs for the specific contracted-for
client facility, explosive events routinely involve collateral damage to adjacent
structures and/or public spaces and also place these additional lives and property
at risk. This is a fact that cannot be ignored. The performance of the client
facility must be considered within a regional context in order for the client and
the design team to exercise the highest standard of industry care and to reduce
the extent of design professional liability. The creation of a highly protected
facility that exclusively focuses on protection of that specific facility’s assets
and does not take into consideration how its performance may affect adjacent
46
DESIGN CONSIDERATIONS
structures, public spaces, and their populations cannot be considered the most
prudent design solution.
Facility Structural System Protection of the facility’s primary and secondary
structure from progressive collapse and disproportionate damage associated with
explosive events external or internal to the building is a baseline protective designconsideration in consequence evaluations. Catastrophic failures of an entire
structure or a significant portion of it generally place large occupant populations
at risk, both as a result of the immediate event effects and of disruption to ERR
operations. Additionally, these structural failures invariably cripple critical business missions, sacrifice institutional operational viability, and adversely affect
restoration of day-to-day operations.
Facility Facade System Explosive events, either internal to or external to the
building, are likely to engage the facade in managing applied forces significantly
in excess of the thresholds associated with conventional wind, seismic, and gravity loading. This routinely results in catastrophic glazing and glazing system
failures, which are historically responsible for significant loss of life and injury
to both occupants within and external to the targeted facility. Airborne glass
fragmentation continues to be identified as one of the most lethal aspects of
building component responses to explosive events.
Facility Evacuation, Rescue, and Recovery Systems While the immediate devastating effects of explosive events can be limited by protective design strategies,
post-event evacuation functions and the subsequent activities of rescue and recovery are dependent upon the integrity of stairs, vertical transportation systems,
power, lighting, voice communication, smoke management, and other systems.
These systems can be placed out of service throughout the building or suffer
disproportionate damage if they are not specifically designed in response to the
project design basis explosive threats. Maintenance of these critical ERR systems, outside of the immediate explosive event zone, is essential to reducing
the extent of resulting lethality, injury, and facility disruption. Design professionals, when evaluating explosive event consequences to ERR systems and
operations, must take into consideration that these systems may be exposed to
supplemental damage due to structural instability. Also, that extraordinary (terrorist) criminal event modalities have historically been strategically planned so
that phased explosive events occur with the purpose of targeting emergency responders and those residual occupants on the scene who have not yet been evacuated.
HVAC Systems Explosive events are traditionally accompanied by the development of significant quantities of airborne particulate contaminants, including
glazing fragmentation, toxic materials, and other byproducts of building component demolition. These are inevitably deposited within the HVAC system as a
result of system breaches associated with the explosive event, or as a result of
CONSEQUENCE MANAGEMENT
47
fans continuing to operate after the explosive event. This contamination can be
of such significance as to warrant the demolition and entire replacement of the
HVAC fan systems (including filters, coils, etc.), fan plenums, and distribution
duct work.
Business Continuity The design professional community should naturally address issues of life safety as a first priority. However, considerations for institutional business continuity cannot be ignored as part of the initial consequence
evaluation process, and may in fact be a client’s most clearly stated objective.
The extent of the business continuity protective design mandate will depend upon
the client’s mission statement and expectations. It will also be dependent upon
the extent of resources available to provide building protective design strategies beyond those required to achieve the consensus of life-safety performance
standards. In general, clients’ expectations regarding the financial benefits
of business continuity are well stated. However, ownership routinely anticipates that the building codes and design professionals’ experiences will adequately guide the design in the protection of occupants on site. This potential
disparity in consequence management expectations must be identified and
addressed accordingly.
Restoration of Operations Explosive events are violent conflagrations with the
potential to significantly disrupt facility physical plant, affect life safety, and
violate viable occupancy. Every explosion will be initially treated by the law
enforcement community as a crime scene, by the building department as a potential condemnation, by underwriters as a time for claim investigations, and
by others as the site of numerous activities other than those originally intended
as part of the building’s use profile. There will be an inevitable period of investigation and severe facility occupant access restriction. Design professionals
should engage the client in post-explosive event consequence evaluations.
These should very specifically focus on the extent of post-event facility demolition and restoration that is tolerable in order to meet client expectations regarding facility condemnation by the local building authority, the
anticipated time and cost of reparations, and the extent of time required to
reestablish baseline business-mission operations.
Target Attractiveness To the extent that crimes involving the use of explosives
can be deterred, a significant advantage can be achieved. While there are no specific predictive mechanisms that can be referenced as an engineering and design
guideline for criminal deterrence, especially for extraordinary (terrorist) attacks,
the concepts of target hardening and inaccessibility are universally accepted by
the law enforcement and intelligence agencies as rational approach strategies for
crime reduction. Consequence evaluations should consider, at all stages of
design, whether the facility will be an appealing target. It is especially important that target attractiveness be considered in the earliest phases of site
selection and facility location within the site. Also, the extent of accessibility
48
DESIGN CONSIDERATIONS
and opportunity for an aggressor to deliver the explosive to the target needs
evaluation. Reduction in the perceived opportunity to achieve both of these potentials is a design goal worthy of substantial consideration.
2.5.2 Function Redundancy
This subsection identifies and provides insight into the benefits associated with
the management of post-explosive event consequences through the application
of redundancy in various building architectural elements, programmatic spaces,
and/or building systems. Incumbent on the discussions regarding redundancy is
the necessity to advance the concept of locational diversity, as adjacent redundancies are not efficient managers of consequences from an explosive event. Simply
stated, adjacent assets are most likely to suffer the same damages.
This subsection text includes the identification of building systems involved
in evacuation, rescue, and recovery, and discusses how redundant ERR systems
may reduce the requirement for structural hardening or may otherwise inform
structural concepts, while reducing consequential effects. This is essential as
ERR systems have a primary role in enhancing facility life safety in response
to an explosive event and the secondary consequences of such a circumstance.
This subsection also addresses business mission-critical systems and spaces
and how redundancy and diversity of these systems may beneficially affect structural and facade designs from the standpoint of explosive event protection. Other
aspects of redundancy associated with building components, systems, or program spaces are also discussed with a focus on how they may inform the structural designs in terms of acceptable post-explosive event effects.
The subject of structural redundancy, for primary and secondary structural
elements, is not covered in this subsection, as it is the specific subject matter of
other chapters in this handbook.
A primary facility performance goal for the design professional is to ensure
that evacuation, rescue, and recovery systems remain operational outside of the
specific area of influence associated with the project-specific explosive design
basis threat event. This is a significant challenge since there are a number of
systems that participate in ERR. Design professionals should anticipate that the
following systems, at a minimum, are considered part of the ERR protection
strategy:
1.
2.
3.
4.
5.
6.
7.
Fire suppression systems, including their water supplies
Fire alarm, detection, exiting, and annunciation systems
Emergency voice communication systems
Emergency lighting systems
Video surveillance systems supporting ERR spaces and operations
Emergency way-finding and exiting systems (other than exit signage)
HVAC chemical, biological, and radiological (CBR) agent control
CONSEQUENCE MANAGEMENT
8.
9.
10.
11.
12.
13.
49
HVAC smoke control, purge, and pressurization systems
Elevators designated for Fire Department use
Access control systems engaged in emergency exiting
Fire and security command and control facilities
Emergency power systems supporting the systems listed above
Emergency exiting stairs and passageways
These are complex systems, with components distributed throughout a building. The opportunity for their disruption is high as a result of the high-energy
devastation characteristic of an explosive event. Consequently, their survival outside of the immediate explosive event zone will depend upon significant investments in hardening of walls and slabs to protect both the front-end monitor and
control portion of these systems and their extensive and pervasive distribution infrastructure. Consequently, this is rarely the most economical approach. It is also
least likely to provide reliability equivalent to that of a design that implements
redundant systems located with geographic diversity, whose extent of segregation is based upon informed explosive event calculations and modeling. In summary, placing critical ERR system components outside of the blast effect
zone is more often a more survivable and robust solution than attempting to
provide a hardening strategy to overcome single points of vulnerability and
consequent failures.
Business-mission continuity depends upon the survival of critical systems,
equipment, information, and intellectual resources. Consequently, they must be
located within safe and secure spaces that are appropriately designed for human
occupancy and provide conditions that do not exceed the systems’ thresholds of
environmental tolerance. These are fragile components, systems, and/or conditions that are as susceptible to disruption or compromise from an explosive event
as the previously discussed ERR systems. Depending upon the type and extent
of service and/or product provided by the institution, there may be a requirement for protecting a particular portion of or, alternatively, the entire facility.
Design professionals, in the Preliminary Planning and Conceptual Design
phases, will need to partner with the client and determine the extent of the
facility that must remain free of explosive event effects (as well as exterior
support infrastructure) in order to achieve the client business continuity
expectations.
If a backup project site protective design approach is not identified as a viable
option (see Section 2.5.4), then the strategy of individual physical and intellectual asset protection will need to be compared to a strategy employing redundant
systems and personnel support within the project facility. These will need to be
located at a dimensional separation established by informed explosive event calculations based upon the project explosive design basis threats and the locations
where these threats can be delivered. These locations will be dependent upon the
facility’s security program. Where local hardening options exist, they should be
50
DESIGN CONSIDERATIONS
carefully examined and a determination made as to whether they are the most
cost-effective and reliable approach to a robust business continuity design concept. If so, then critical business operations and their support systems can be
confined to a compact and unified occupancy program within a locally hardened
portion of the structure.
External explosive events may place significant portions of a building’s facade at risk, and the failure mode may involve large portions of the exterior
envelope. Based on the explosive event specifics, including the threat assessment
explosive DBTs, standoff, and architectural envelope design criteria, post-event
facade response may involve significant inbound facade debris that could negatively impact life safety, ERR operations, and business continuity programs.
Design professionals, in the Preliminary Planning and Conceptual Design
phases, should partner with the architectural team and develop preliminary
ERR design strategies, informed by facade explosive event experience, so
that determinations regarding facade hardening versus redundant diverse
ERR systems are given due consideration and the proper balanced solution
selected.
While ERR components and operations and business continuity mission expectations traditionally receive the majority of design focus for protection from
an explosive event, client requirements may dictate that other physical, intellectual, or operational assets survive the DBTs. To address this additional requirement, design professionals will need to clearly understand the type and locations of assets at risk. Subsequently, the post-explosive event architectural,
structural, mechanical, electrical, and other related systems’ performance requirements will need to be defined in a manner responsive to the asset protection
profile. Based on the complexity of the necessary support infrastructure, design professionals should conceptually investigate the option of individual
system hardening versus redundant support systems at geographically diverse locations outside of the calculated explosive event effect zone. Even if
a redundant diverse support system is selected, it should be recognized that local hardening will likely be required at points of connectivity to the protected
space(s).
In some cases, functional redundancy is not an option. The preservation of
unique cultural artifacts, the protection of unique financial assets, co-located intellectual resources, critical unique industrial or financial processes, and other
single-source assets do not allow the design professional the benefit of protection through diverse repetition. These will require competent hardening strategies specifically created in a manner responsive to each asset’s vulnerability.
A discussion of functional redundancy and diversity, and mitigation strategies involving an analysis of hardening versus repetition and separation is not
complete without consideration of effective strategies to preclude the introduction of explosives to the facility or to assets at risk, through the implementation of a competent security program. Design professionals should vigorously
pursue with ownership opportunities for threat reduction. The potential for
implementing this approach will yield a more secure facility with greater
CONSEQUENCE MANAGEMENT
51
functional capability for ERR and the least threat to business continuity or
other single-source, high-value assets.
2.5.3 Building Location
This section explores and defines the opportunities that the structural and blast
consultant engineers have available to assist the client in understanding the structural, architectural, site civil, constructability, cost, schedule, and operational impacts associated with explosive event protection during the site selection process.
Design professionals are generally given the opportunity to assist a client in
site selection, or, alternatively, are asked to develop the best structural protective
design response for a building on a particular client preselected site.
Site selection criteria will be influenced by the threat assessment as identified
in Section 2.4 of this handbook.
When design professionals are engaged to work with the client on identifying the suitability of a site, or sites, and the placement of a structure on
those site(s), the decision should be informed by a competent and thorough
threat and risk assessment. This assessment must clearly address the qualitative
explosive threats (vehicle-borne, man-portable/satchel, maritime vessel– delivered, airborne delivery, standoff military ordnance, etc.), the quantitative design
basis threat values (amount of TNT equivalent for each of the threats identified,
shaped or unshaped charge, etc.), and the extent of opportunity for establishing
standoff between the threats and the protected facility. These are discussed in
more detail in Section 2.4.
Similarly, when clients request that design professionals work within a given
predetermined site (i.e., when informing the site selection process is not part of
the design professionals’ responsibilities), the limitations which that specific site
will impose upon the structural engineers and other design professionals should
be identified. This should be based upon the risk assessment results associated
with an assessment performed for that specific site.
Site selection and building placement locational consequences can be addressed through a competent process of explosive event analysis. This analysis
must be informed by the threat assessment. It should also be informed by standoff distance determinations, the evaluation of the resulting extraordinary loads
imposed upon the facility by the explosive event threat, and the opportunities for
mitigating strategy protective design concepts available to the design team.
Elimination of the explosive event threat remains the most attractive,
competent, and cost-effective process in developing designs for facilities on
any site. However, based on the threat assessment, it may be unreasonable to
assume that the facility will not be exposed to explosive event effects. Maintenance of standoff between the explosive device and the asset remains as one
of the most powerful mitigating strategies available to the design team for
the development of a structure that affords the best opportunities for risk
and consequence management. In the case of standoff military weaponry (such
as shoulder-fired ordnance), should it be identified in the threat assessment as a
52
DESIGN CONSIDERATIONS
design basis threat, standoff distance will likely be less important than obscuring line of fire, the provision of a predetonation screen, or the provision of other
measures, as this form of attack can take place from a considerable distance.
In general, applied pressures and the duration of their application (impulse) to
the structure drop off exponentially as the distance between the protected structure and the explosive event increases. The protection of the facility’s facade;
primary and secondary structure; evacuation, rescue, and recovery systems; business continuity–related intellectual assets and physical plant; and other preidentified entities worthy of protection will require a high level of design sophistication. The design process will need to balance budget allocations and architectural
and structural design responses based upon the achievable standoff between the
assets to be protected and the explosive event delivery location.
As part of the site selection process, the blast consultants, working with
the threat assessment identified DBTs, should work closely with the client
and design professional team. This process should establish the pressures
and impulses that will be applied to the structure, its facade, inbound utilities, and other critical site civil features, based on the maximum achievable
and realistic standoffs. This is essential so that an initial determination can be
made on how these extraordinary loads (in excess of gravity, wind, and seismic) will affect the protective design capabilities of the following major building
systems and/or afford protection to the following building components and occupants:
1. Major/critical inbound utilities, such as fire protection water service, utility
electrical power, information technology and data services, domestic water,
steam, and gas
2. Facility access control points for vehicular entry
3. Site-located critical utilities such as transformers, generators, cooling towers, fuel storage, domestic and fire service water storage, and communications infrastructure
4. Facility facade
5. Facility primary and secondary structural systems and elements
6. Vertical transportation systems, including stairs, escalators, and elevators
internal to the building (these may be exposed to explosive event effects
should the robustness of the facade design be less than the DBT event
calculated pressures and impulse loads)
7. Evacuation, rescue, and recovery systems as identified previously, in Section 2.5.2, internal and external to the building (these may be exposed to
explosive event effects should the robustness of the facade design be less
than the DBT event calculated pressures and impulse loads)
8. Occupants within the facility
9. Business mission-critical systems and equipment located within and/or external to the facility
CONSEQUENCE MANAGEMENT
53
The opportunity to achieve maintenance of standoff between the explosive
DBTs and the project assets, as noted above and as supplemented for each specific project, should be one of the primary focuses of the site selection process.
This should involve a careful evaluation of how the explosive device delivery
threat, as identified in the threat assessment, can be competently separated from
the asset(s). Rational solutions involving vehicle anti-ram perimeters, pedestrian
interdiction barriers, and other site civil and landscape features should be considered for their implementation feasibility. Inability to institute these protective
features may have consequences on the building’s design. Consequently, the following strategies should be considered:
1. To increase survivability of critical inbound utilities, and, therefore, the reliability of ERR systems and operations, business continuity, and the mitigation of secondary event effects (such as the ignition of a gas main as a
result of the explosive event), these utilities should be located outside of
an explosive event delivery zone. Alternatively, they should be provided
with hardening protection and redundant diverse routing at dimensional
separations identified by the blast engineer. Utility exposure and explosive event survivability evaluation is an initial critical design determination, as it may inform substantial work and expenditure for utility
hardening protection or diverse routing and/or establishing long lead
time approvals and/or feasibility studies by local utilities.
2. To increase protection for building occupants, ERR systems, critical
business-mission operations, and other predetermined internal assets, the
facade system for the building should be reviewed for its ability to be designed to reduce the threat of airborne fragmentation to both internal building occupants and those external to the building. This should include a determination as to whether a design can be implemented whose post-event
performance will be adequately robust to preclude additional protective design measures internal to the building, or whether the failure mode requires
supplemental protection of internal and external assets. Facade robustness
evaluation is a critical initial design determination, as it will influence
the building’s architectural aesthetic, internal space program assignments, and the dedication of additional expenditures for hardening
and/or diverse redundancy of ERR and business mission-critical systems.
3. Initial structural design concepts should be evaluated for their ability to
increase the survivability of the primary and secondary structure, and, consequently provide significant enhancements in life safety and businessmission continuity, and reduction in both decontamination and restoration
time to resume normal operations. These conceptual structural systems
should be reviewed for their ability to be resistant to progressive collapse
and disproportionate damage, based on the building location and the available standoffs. Structural system survivability is a critical initial design
54
DESIGN CONSIDERATIONS
determination, as the type of structural system for the building may,
in fact, be determined by explosive event management, in lieu of being
conceived to address traditional gravity, seismic, and wind loadings,
in order to achieve the post-explosive event performance results. If a
conventional structural system cannot satisfy the security performance requirements, then site selection and/or building placement on the site may
have blast protection requirements that inform member sizing, connections, standoff, and other major portions of the design with consequential
effects on construction schedule, budget, and the architectural program.
4. To ensure the most competent and effective means to facilitate evacuation, rescue, and recovery operations, the designers should evaluate egress,
emergency responder vehicular access, and emergency responder entry.
This should be performed through a process of establishing their exposure to the project explosive DBTs and the separations/standoffs that can
be provided by the particular site. Vehicular access points to the site, approach configurations to the facility, and the conceptual location of exit
stairs and elevators should all be reviewed for survivability. The survivability of ERR facility components is a critical design determination,
as it may inform major aspects of the site civil and landscape plan, and
the location, quantity, and type of the facility’s vertical transportation
systems, as well as the hardening required for these systems.
2.5.4 Building Dispersal/Distribution of Functional Programs
This subsection addresses the opportunities for consequence evaluation and management associated with the physical separation of internal building program
spaces, critical ERR and business continuity support systems, and physical assets and occupants. It focuses on the relationships among facility vulnerability,
facility diversity (geographically diverse facility entities and, also, redundancy
and diversity of internal building assets at a single site), the blast analysis process and the associated analytic results, and how these factors and engineering
analytics can inform whether physical separation and the resulting diversity are
a more rational protective design solution than co-location and structural, architectural, and ERR system hardening.
The threat assessment process discussed earlier identifies credible threats appropriate for mitigation. If explosive event threats are subsequently identified,
then both the site selection and facility design process will need to address their
vulnerability to the specific explosive event types, design basis threat quantification, the means of delivery, and their adjacency to the protected asset.
Vulnerabilities will occur wherever the explosive event threat can be delivered at distances determined by the blast consultant to impose pressures, impulses, soil movements, and other explosive event effects in excess of the loads
prescribed for management by the building code. Additionally, vulnerabilities
may exist where these explosive events create conditions in excess of the criteria
CONSEQUENCE MANAGEMENT
55
stipulated in the client’s published design standards for project performance. It
is essential for a qualified blast consultant to inform the design team of these
extraordinary loads, vibrations, soil displacements, and other consequential
effects, and the distances at which they will be applied to the facility’s structure, facade, and critical systems. These values should not simply be obtained
from resource and reference materials unless they are reviewed and subsequently
modified and/or approved by a qualified explosives and structures subject matter
expert with extensive experience and access to government-approved computer
modeling software.
Subsequent to the identification of the anticipated explosive event effects,
based on the type and delivery location opportunities for the explosive threat,
conceptual building placements on the site and conceptual identification of ERR
and business-mission or other critical assets should be identified by the design
professionals working with the client team. Various options are likely to exist
in the Conceptual and Schematic design phases, and each of these should be reviewed to identify explosive event vulnerability exposures. Based on the agreedto, post-explosive-event facility design survivability performance criteria, the design team should consider:
1. The development of maximized enforceable standoffs.
2. The utilization of protective design facade, structure (primary and secondary), and internal partition hardening.
3. The addition of redundant ERR and other systems located diversely enough
within the facility to ensure survivability to a single or multiple explosive
events, as defined in the threat assessment. Note that this must take into
account both external and internal explosive threat modalities.
4. The repetition of those portions of the facility at a diverse site (essentially
another standalone clone of the assets at risk), at a distance ensured to be
free from the explosive event effects at the project site as determined by
the blast consultant.
Life-safety concerns for facility occupants and those external to the building are not specifically addressed by the concept of redundancy and/or diversity,
other than through the benefit they receive from the protection afforded by survivable evacuation, rescue, and recovery systems. Those within the immediate
blast effect zone (an area of influence that generally can be well estimated by
the blast consultant) will suffer traumatically unless they are also protected by
blast-hardened elements and free from exposure to airborne debris. However,
the protection of those systems that aid ERR operations will significantly enhance the life safety of those who will need to exit the facility and provide rapid
and efficient access by emergency responders immediately after the explosive
event. These activities are essential in assisting those facility occupants affected
by the event or those building occupants who fall within the ADA-defined profiles and require supplemental assistance for evacuation. In summary, diversity
56
DESIGN CONSIDERATIONS
and redundancy are strategies best employed for the protection of ERR systems
and building features, critical business-mission systems, equipment, and operations or other client-defined physical assets. The initial phases of consequence
evaluation and management should occur in the Conceptual and Schematic
design phases and include a review of the opportunities for redundancy and
diversity and the use of hardened facility design elements.
The above notwithstanding, certain facility operations and client requirements
may dictate the duplication of staff either within a facility or at a remote site.
This will occur when individuals are considered to be of such importance that
they are an integral component of business-mission continuity support. When
this is a design parameter, the structural engineer, architect, and blast consultant must address the project’s explosive design basis threats in a manner
that identifies the extent of segregation required between the area where those
threats may occur and the proposed location for this supplemental protected
population.
This process must evolve from the development of initial “threat maps,” which
are based upon the available architectural concept drawings. These maps should
be created to identify the location of explosive event delivery opportunities, both
external and internal to the facility. This will require the parallel conceptual development of a facility security operations program. Consequently, a security
professional should be part of the design team’s composition so that security
systems and management operations are designed to limit the potential size,
delivery strategy, and locational placement of explosives.
Once there are reasonable key design assumptions regarding explosive event
threat delivery locations, explosive event analytics should be performed to determine whether a co-located population of duplicate staff (and the infrastructure
required to support them) can be adequately protected within a single building,
or whether they need to be physically relocated to a remote site. Although this
may appear to be a daunting and challenging process during the early stages of
the project, where designs are truly conceptual, this will provide the best opportunity for informing the major structural and architectural aspects of the design. The longer the process of determining how staff will be protected is
delayed, the more difficult the protective design solutions will become. This
will therefore generate a compromised set of strategies that will cost more
and provide less competent protection.
2.5.5 Disaster Recovery and Contingency Planning
Disaster recovery and contingency planning provide carefully conceived and
tested alternatives for taking action when things go wrong. Since no privatesector structure can be built to survive every explosive attack, and since no security plan is perfect, every organization that believes there is a threat to its facility or proximate environs should create disaster recovery strategies, contingency
procedures, and business recovery plans. Indeed, most boards of directors and
insurance companies demand no less.
THREAT REDUCTION
57
The purpose of contingency planning is to protect building occupants and
speed business resumption, including regaining access to data, communications,
workspace, building support systems, and other business processes. In short, the
consequence management and the other security design concepts discussed in
this handbook can and should support business continuity planning. Early in the
conceptual design process, therefore, the design team should discuss post-blast
recovery with ownership in the specific context of their contingency planning
expectations. Design professionals should insist on reviewing with the client
the opportunity to incorporate contingency planning and business recovery
concepts into their building design.
From the occupants’ perspective, the single most significant reason why contingency planning is important is that lives could hang in the balance. To the
eventual building client institution, effective business recovery strategies are also
critical in maintaining the public’s confidence in the enterprise during a crisis.
2.6 THREAT REDUCTION
The design professional, security, and law enforcement communities acknowledge that facilities can be safer environments when those threats that place the
institution and its occupants at risk are managed by a process of threat reduction and/or, preferably, elimination. When the law enforcement and intelligence
agencies cannot ensure management of exposures to an explosive event, then institutions can achieve supplemental threat reduction by being placed as far from
the threat as possible, so that the residual effects of the threat at the maximum
achievable standoff distances are then addressed by the facility’s site selection,
facility placement within the site, and the facility’s design. These relationships
will be discussed in this section.
Explosive events can be categorized as those that occur:
1. As a result of the purposeful, beneficial, and legitimate use of explosives
(e.g., rock excavation for foundations)
2. As a result of purposeful, willful, and malicious criminal acts, using explosive materials (e.g., a terrorist bombing)
3. As a result of a criminal act that does not initially use an explosive material,
but that subsequently engages other resident agents possessing explosive
capabilities (e.g., an arson event that then engages a compressed gas vessel)
4. As a result of an accidentally occurring event (e.g., a high-pressure steam
line rupture)
5. As a result of a naturally occurring event (e.g., a lightning strike engaging
an electrical transformer)
The law enforcement and intelligence communities and the security professionals responsible for securing a city, neighborhood, and/or facility site face an
58
DESIGN CONSIDERATIONS
impossible challenge when tasked with the complete elimination of the explosive
event threat. The prevalence of the suicide bomber creates exceptional opportunities for explosive event delivery that preclude providing the design professionals
with a design basis threat of “no explosives.” Consequently, when design professionals are tasked with the development of protective design strategies
for a site, building, and its occupants, they must insist on obtaining the specific explosive design basis threats, which are identified within a competent
threat and risk assessment created with contemporary input from the law
enforcement and intelligence agencies on explosive material types, delivery
modalities, and any other related available specifics. This information mining
is essential if credible structural and architectural facility designs are to provide
real-world protection.
Furthermore, it should be recognized that the law enforcement and intelligence communities cannot effectively identify the extent of explosive event
threat reduction that a facility security program will accomplish, unless, perhaps, it is a government facility for which they have direct management control
and involvement. In most cases, on-site and internal-to-building explosive event
threat reduction must be identified by the client and their security professionals, or a highly qualified security consultant. This information must also inform
the project threat and risk assessment so that quantifiable explosive design basis threats are represented on the previously discussed facility threat maps. This
process will allow a clear and graphic representation of how explosive threat
reductions can occur, starting at the facility site perimeter and migrating to the
most sensitive internal2 areas requiring protection. The intent of threat mapping is to clearly identify explosive quantity and locational placements to
inform the design professionals on how the explosive event DBTs are reduced and/or eliminated as a result of being subjected to concentric layers of competently applied security measures. This is the information which,
as represented on the threat maps, will inform how threat reduction can be
used by the design team to address malicious criminal acts involving the use of
explosives.
To reduce the threat associated with naturally occurring events, the designers should gather relevant, project specific information and review the
site and the facility designs at the earliest possible stages of design. The purpose of these investigations is to identify those sources of explosive event opportunity worthy of supplemental consideration and protection. This should involve
the identification of other explosive event substances that contain inherent highenergy sources such as:
1.
2.
3.
4.
5.
Compressed gas vessels and/or piping
Liquefied natural gas vessels and/or piping
Fuel storage and piping other than gaseous agents
High-pressure steam vessels and/or piping
Stored munitions and materials used in their fabrication
THREAT REDUCTION
6.
7.
8.
9.
10.
59
Electrical transformers
Combustible airborne particulates
Combustible chemicals and agents other than explosives
Chillers, boilers, and furnaces
Numerous others
These may occur either adjacent to or resident within the site and/or facility,
or exposures may occur as a result of person and/or vehicle (motor vehicle, rail
transit, aircraft, or maritime vessel) delivery within or adjacent to the facility.
2.6.1 Accidental Explosions
This subsection identifies potential accidental explosions that should be taken
into consideration by the design professional team. The intent is to ensure that
design professionals—especially the structural, mechanical, and electrical engineers and the architect—are aware of and provide protection for explosive events
other than purposeful, malevolent criminal activity.
Explosive events occur as a result of the release of high-energy sources. While
traditionally this condition is conceived of as the result of ignition of an accelerant, such as highly flammable liquids, compressed gases, or explosives, other
high-energy sources—such as steam or water under extreme pressure, chillers,
boilers, and other systems routinely found in facilities—are also candidates for
imparting extraordinary pressures over extremely short durations.
The unabridged edition of the Random House Dictionary of the English Language defines an explosive event as “the act or an instance of exploding; a violent
expansion or bursting. . .” and “a sudden, rapid, or great increase.” While the
pressures, impulses, and plasma effects associated with an explosive event
using extremely high-energy substances such as C4, Semtex, RDX, TNT, or
their equivalents, are regarded as upper-level design thresholds for protection, design professionals should carefully review the entire project design
with the intention of identifying other building systems and/or programmatic spaces allocated to the storage and/or distribution of substances and
materials that could participate in an explosive event. These systems and areas of programmatic use are unlikely to be identified by a traditional security
threat and risk assessment, as these activities usually focus upon the purposeful
introduction of high-energy explosives as the anticipated design basis threat.
At the earliest possible design phase, the structural engineer, mechanical and
electrical engineers, and architect should review the facility design and identify
whether any of the systems, equipment, or environmental conditions identified
previously are resident within the project scope. This process may involve making specific inquiries of the client, based on the sophistication of their occupancy
program, regarding what explosive exposures they know exist, as a result of their
detailed familiarity with the facility operations that will occur in the project. The
design team should list, as part of the protective design strategy’s key design
60
DESIGN CONSIDERATIONS
assumptions, the potentially explosive operations, systems, and/or materials
and substances that will be present as part of the building’s use.
A list of potential explosive event conditions should be created based on the
aforementioned plan review and client inquiry. Individual mitigating strategies
should be proposed by the design team to address these exposures. This may
include the following:
1. Elimination and/or reduction of explosive sources and/or quantities from
the project
2. The introduction of detection systems to identify the conditions associated
with the development of an explosive event condition prior to its detonation
and/or release point
3. The introduction of suppression systems capable of reducing the postexplosive-event effects
4. The introduction of systems that extract the explosive event agent, upon its
detection, and prior to its detonation/ignition
5. The elimination of ignition sources
6. Management of access control, for both persons and vehicles, to preclude
entry and accidental initiation of the explosive event
7. The introduction of explosive event pressure relief features that direct the
effects of the explosive event in a manner that provides acceptable postevent effects
8. Informed placement of potential explosive systems and/or environments so
that their post-event effects meet the project’s security performance criteria
9. The use of appropriate signage and graphics to prevent occupancies, activities, and/or behavior that could trigger the event
10. The use of hardening to preclude unacceptable and/or disproportionate
damage from the event
These and other mitigating strategies should be considered for implementation
by the design team to enhance the facility protection profile.
2.6.2 Intentional Explosions
This subsection addresses purposeful extraordinary criminal activity associated
with the use of explosives as a means of attacking persons and property. Refer
to Section 2.3, “A Brief History of Recent Terrorist Attacks,” which identifies
sources of related explosive event history and demographic information. While
engineering design guideline publications can assist the design professional by
providing pertinent related historical information, ownership and design professionals must anticipate that history will repeat itself and that the design and
client community have a responsibility to engage in predictive security planning.
THREAT REDUCTION
61
This handbook advocates the use of the extensive explosive event information available from numerous agencies, such as the FBI, the Department
of State, the Department of Defense, the Department of Homeland Security, the U.S. Armed Forces, and others, as a means to reinforce confidence
in providing—well beyond the publication date of this manual—protection
against explosive events.
While the primary use of explosives was in the military arena, criminals and
terrorists have adopted the use of explosive devices as the most commonly used
tactic to inflict devastating results. Contemporary with the publication of this
handbook, law enforcement agencies, the intelligence community, terrorism experts, and security professionals continue to emphasize that the manportable improvised explosive device (MPIED) and vehicle-borne improvised explosive device (VBIED) are the weapons of choice for the criminal
to use when contemplating an attack against a population and/or facility,
with the intention of achieving mass loss of life, disproportionate property
damage, and the inevitable publicity this generates.
Facility designs outside of North America are also subject to the use of explosives as part of the criminal/terrorist act modality. In addition to the MPIED and
VBIED delivery methods, the use of standoff military ordnance, especially the
use of shoulder-fired arms (such as the rocket-propelled grenade [RPG]) remains
a significant concern and should be given careful consideration by a structural engineer and architect as part of a competent protective design mitigating strategy.
It is especially important, for work outside of North America, that design
professionals insist upon the use of a competent threat and risk assessment
process that is specific to the project facility, in order to determine whether
standoff military weaponry is to be considered as part of the mitigation
criteria.
It is also essential for the design professional team, especially the structural
engineer and architect, to understand the relationship among procurement of explosives, their delivery, placement, and subsequent detonation, in order to clearly
understand the rational and design consequences associated with the identification of design basis threats. The protection strategies and eventual design implementation concepts and components of a facility’s structure, facade, ERR systems, and business-mission continuity are founded upon mitigation of either a
predetermined explosive event type, size, and detonation location, or a predetermined pressure, impulse, and post-event facility performance standard. Consequently, the establishing of quantifiable design basis threats is a fundamental first step in providing the necessary guidance to the design team, as this
will be the basis for subsequent analysis and design mitigation development.
Explosive design basis threats (DBTs) must be provided as part of the threat
and risk assessment by a competent security professional. The project-specific
and identified DBTs will be based upon the opportunity for the aggressor to procure the explosive, the ability for it to be delivered to the point of attack, and
the opportunity to achieve a detonation sequence that will achieve the full value
of the DBT value. Threat and risk assessment professionals, working with the
62
DESIGN CONSIDERATIONS
law enforcement and intelligence communities, have resources to assist in the
development of DBT values based on these generic parameters. However, it is
essential for the design professional team to participate in the risk assessment
process by the provision of project-specific information, at the earliest stages
when conceptual designs are adequately advanced. This is necessary as it will
inform the ability for explosive delivery mechanisms to be considered credible or
not credible, based on project design features that affect the mobility of vehicles
and persons intent on delivering the VBIED and MPIED, respectively. While
the threat and risk assessment authors can provide quality information regional to the site, the extent to which the project site and facility physical
plant are at risk from specific explosive event DBTs must be informed by an
interactive process of information exchange between the threat assessment
and design team professionals.
Large quantities of explosives are readily delivered by vehicles. These values are considered public sector information-sensitive. However, there is a trend
toward this information becoming common knowledge within the design community as more explosive event protection mitigation is practiced. However, this
handbook continues to respect the need for this information to remain confidential. Consequently, the explosive delivery capabilities for passenger-operated
vehicles (cars); passenger vans, sport utility vehicles, and pickup trucks; small
box trucks; large box trucks and trucks with vessels carrying liquids; and the
larger classes of vehicles such as tractor trailers have all been assigned explosive
event design basis threats by the law enforcement community. Quality sources
of VBIED DBT information are the ATF, TSWG (Technical Support Working Group), and other DHS agencies. Design professionals should request
this information from the project’s security professionals engaged in the
project threat and risk assessment.
MPIED DBTs have been credibly established as a result of their extensive use
in numerous criminal acts. Again, there is a range of explosive event DBT sizing
based on the delivery mechanism. Individuals may deliver explosives either by
transporting them directly as a body-worn device or as a separately carried parcel
or satchel. The increased use of well-packaged high explosives allows for competent concealment and provides the law enforcement and security professional
with significant challenges for rapid identification through the routine use of informed visual observations. There are emerging programs that employ behavioral identification, parcel recognition, and other predictive strategies employed
under the nomenclature “situational awareness.” These are supplemental procedures only; they are helpful in explosive device identification, but should not be
considered substitutes for physical protective designs.
The body-worn explosive device DBT is not insignificant and is capable
of inflicting structural damage. The extent of this damage can be dramatically increased through the use of high-energy explosives and specific charge
shaping.
The satchel may be carried or packaged in a wheeled parcel pulled by the aggressor. Wheeled parcels often are harder to detect because the ease with which
VULNERABILITY REDUCTION
63
they can be transported reveals less through body language. Consequently, the
MPIED has a potentially large value range. It is significantly in excess of the
body-worn DBT value.
Explosives can also be purposefully delivered using the U.S. Postal Service,
express couriers, common freight carriers, or other similar second- and thirdparty delivery agencies. These differ from the MPIED threat, as they arrive
cloaked by the legitimacy associated with an expected delivery mechanism and
devoid of the opportunity for profiling the carrier. The range of DBT value is
entirely dependent upon the parcel/freight weight and size management threshold criteria established by the carrier. Suffice it to say that the range of explosives that can be delivered by a carrier generally exceeds that which an aggressor
could deliver personally. With the advent of GPS and other handheld intelligent
geospatial and initiating technologies, criminals are capable of remotely detonating shipped explosives at explicit latitudes, longitudes, and elevations, allowing
targeting by building, by floor, and by room, assuming satellite signal transmission strength is adequate.
In summary, design professionals should use the security consultant’s estimates of the VBIED and MPIED threats in the design of protection strategies for the structure, facade, ERR systems, and so forth. They should insist
upon active participation in those portions of the threat and risk assessment
that involve the identification of building design features and components
that can affect the type, size, and location of a bomb. Also, a competent security systems and operations professional should work with the threat and
risk assessment and design team to advise on the method to limit the delivery of explosives.
2.7 VULNERABILITY REDUCTION
This section addresses the critical topic of how vulnerability of an institutional facility, including its physical plant (buildings), occupants, and assets
can be secured—both from the standpoint of safety and of business-mission
continuity—by the application of rational and justifiable mitigating strategies.
Facility vulnerability can be defined as the capability of, or susceptibility to, being compromised, damaged, destroyed, or otherwise violated in a manner that is
detrimental to human safety and institutional viability.
It is unreasonable to expect that government agencies can entirely eliminate
the threat of a bomb. As mentioned previously and as so aptly demonstrated
during the Second World War, the kamikaze attack or the actions of the suicide
bomber generally are successful, as they defy the basic concepts of military engagement, law enforcement, and protective designs, which assume that even the
enemy or the criminal exercises the intent of self-preservation. In an environment in which criminal threats cannot be dismissed, only diminished, institutions remain vulnerable, and protective designs that address those vulnerabilities are necessary.
64
DESIGN CONSIDERATIONS
Previous portions of this handbook also identified the potential for accidental
explosive events, which also place structures and their occupants at risk. While
these may be significantly diminished by competent planning and protective
design strategies, there always remains the opportunity for exposures resulting
from protective designs that are overwhelmed by larger threats and/or combinations of events that were not anticipated in the risk assessment.
Factors of safety (FOS) are routinely employed in the designs executed by engineering professionals, especially structural engineering, which benefits from
the values that have established industry standards for seismic, wind, and gravity loads. Unfortunately, there are no specific FOSs for intentional explosions,
and for accidental explosions there are very limited published guidelines (such
as those published by electrical utility companies for transformers, by steam utility providers for high-pressure steam distribution systems, etc.) identifying the
quantified extent of additional protection that should be designed for by the structural engineer. Consequently, the identification of facility vulnerability to an
explosive event, which in large part depends on the value of the explosive
threat to be mitigated, remains an inexact science. While the military has developed practical approaches to address military attack ordnance of particular
sizes and types, this information is not explicitly helpful in identifying the vulnerability of conventional structures to explosions associated with nonmilitary
ordnance.
The process of vulnerability reduction must therefore occur within a design
environment in which there is a presumption that the threat cannot be completely
eliminated, that its quantification is an informed but still subjective value, and
that there are no preestablished industry factors of safety. Irrespective of the
difficulty in specifically quantifying explosive events, there remain viable
opportunities and design considerations for use by design professionals to
address facility vulnerability. They are addressed in the following sections
and can be summarized as standoff distance, physical security and protective design, operational security, and structural design responses.
2.7.1 Standoff Distance
Standoff distance is the amount of space denied to an aggressor between his
threat device and the area being protected; for example, ram-resistant vehicle
barriers set 100 feet from the building facade create a 100-foot standoff. Since
the destructive effects of an explosion decrease rapidly over distance, space is
the most effective and cost-efficient safety measure of all the blast-mitigation
strategies available to building designers and engineers. (See the later chapters
on blast phenomena and blast loading for detail.)
The DoD and other federal entities7 provide uniform tables of minimum
standoff distances that the design team can use to establish and enforce their
7
See the National Counterterrorism Center (NCTC) Bomb Threat Standoff Distances table at
www.nctc.gov/docs/2006 calendar bomb stand chart.pdf.
VULNERABILITY REDUCTION
65
agreed-upon security and safety levels. Determining the correct standoff margin with precision is challenging, however, since estimating the required
standoff distance involves an accurate approximation of the bomb’s composition and explosive weight. Because the terrorists hold the initiative, such
calculations are often simply educated guesses and hence become dependent
on the accuracy of the threat assessment.
Employing standoff design principles to protect the facility’s facade and primary and secondary structure, ERR systems, business-continuity-related intellectual property, physical plant, and other protected assets requires sophisticated
planning, additional budget allocations, and probably architectural innovation or
restraint. Putting space between the explosive device and the assets remains
one of the design team’s best mitigating strategies for risk and consequence
management, so realizing the positive benefits of standoff distance should be
a primary focus of the site selection process. Consequently, vehicle anti-ram
perimeters, pedestrian interdiction barriers, and other site civil and landscape
features should be given serious consideration.
2.7.2 Physical Security
Physical security equipment and operational security measures such as personnel, policies, and procedures are proven defenses that can keep explosives away
from the facility. Although security programs are a clear and observable protective measure—think of controlled underground parking and the badges now required in most office buildings—design teams often neglect to build competent
and enforceable physical security equipment and operational security program
spaces into the facility’s infrastructure.
2.7.3 Operational Security
Architects’ and engineers’ failures to anticipate security program needs seem
rooted in their clients’ tendency to institute these measures only after building
occupants begin arriving. But any facility constructed to resist explosive attacks
needs—to cite only a few examples—controlled access areas, delineated zones of
control, and places where vehicular and pedestrian entry is strictly managed. Design professionals should vigorously pursue with ownership the opportunity
for positioning the required protective equipment and security operations
spaces into their planning cycle in a manner compatible with the intended
facility operation’s plan. The final design will yield a more secure facility with
greater functional capability for detection, deterrence, prevention, and recovery.
2.7.4 Structural Design
This subsection provides a nontechnical discussion of how structural design
should be considered as one of the major mitigating strategies for reducing facility vulnerability. Since the analytical and engineering specifics of this topic
66
DESIGN CONSIDERATIONS
are addressed in much greater detail in other sections later on in this handbook,
this subsection is intended to introduce and validate the role of explosive-eventresponsive structural designs. This section also identifies the significance associated with making the determination that structural robustness is to be part of the
facility’s protective design strategy, along with the accompanying effects on the
design process, product, construction cost, and risk acceptance.
The decision to enhance a facility’s structural design to resist the effects of explosive events and thereby reduce its vulnerability will have project management
design implications. These include:
1.
2.
3.
4.
5.
Owner decisions regarding risk tolerance and acceptance
Project cost allocations
The effects on the role of structure in architectural design
Design space implications for architectural use programs
Design implications for mechanical, electrical, and plumbing (MEP) and
other building infrastructure systems
6. Implications associated with construction quality control for both structural shop drawing submittals and during construction field inspections
7. The requirements imposed on document production and management to
maintain confidentiality of the security-sensitive information included in
the construction document packages
Assuming that standoff distances are optimized through a competent security
program, which achieves maximum available separation between the explosive
DBT and the structure, and that a combination of physical and operational security strategies achieves credible reductions to agreed-to explosive DBTs, remaining vulnerabilities from the explosive threat will need to be managed by
the facility’s structure. Facility protective designs are dependent upon reducing project-specific vulnerabilities, and the extent of mitigation that must
be performed by the structure is the residual threat not capable of being
addressed by separation from the threat and the extent of threat that is
not capable of being detected and precluded from access to the site and the
facility.
Even small explosive events create extraordinary load magnitudes in excess of
wind loads on facility facades. These applied loads to the facade must be transferred back to and managed by the building’s structural frame. Additionally, explosive events, based on the project-assigned DBT values, will very likely apply
significant loads directly to both primary and secondary structural members and
elements, such as columns, girders, trusses, shear walls, connections, and slabs.
This may come about as a result of larger DBT values at some distance from the
structure (externally), or these larger DBTs may have proximal/adjacent access
to the structure itself. Structural exposure to and the consequent vulnerability to
these threats can occur from either the VBIED or MPIED, as noted previously.
VULNERABILITY REDUCTION
67
Additionally, these same VPIED and/or MPIED threats may internally threaten
the structure, depending on the project occupancies and the extent and efficacy of
the facility’s security programs in managing the migration of the threat. In summary, the structural frame/load-carrying system of the building will likely
provide the most significant protective value in reducing vulnerability to an
explosive event, as the loss of primary and secondary structure will usually create circumstances that place significant numbers of building occupants; evacuation, rescue, and recovery systems; and business missions at
severe risk.
However, with the decision to engage the structure in facility vulnerability reduction, one of the most expensive building components is now subject to modification in order to provide performance beyond that required by most building codes and conventional wind, gravity, and seismic loading. While European
building codes (such as those in the United Kingdom) currently embrace the requirement for design professionals to address the removal of a single column
(for unspecified reasons, essentially requiring the designers to address a “threatindependent” vulnerability), this requirement has not yet been uniformly adopted
by building codes throughout North America. And, when this requirement has
been included in a particular regional building code, or as a programmatic design
conformance requirement, the removal of a single column may not provide the
extent of protection required to preclude progressive collapse and disproportionate damage from threats of specific sizes as identified within the project threat
and risk assessment DBTs. These DBTs (used in “threat-specific” or “threatdependent” designs) may be of such a size and capable of achieving adjacency
to the structure so that they affect the load-carrying capability of more than a
single primary element. These circumstances, therefore, require a greater extent of mitigation and consequent vulnerability reduction than would otherwise
be provided by the use of the threat-independent criteria only. Consequently,
a fundamental risk management decision must be made by project ownership and codified in the key programmatic design and facility performance
assumptions to be used by the design professionals: specifically, whether the
structural system for the building will provide protection against the actual
project threat assessment identified explosive DBTs (“threat-dependent” design approach), or the extent of protection afforded by addressing the removal of a single primary structural element (“threat-independent” design
approach), or, potentially, both criteria. The client-defined extent of postexplosive-event facility performance must be clearly understood and agreed
to in order for there to be a clear expectation regarding structural performance in vulnerability reduction. This may have a significant project cost
impact, and, during the initial Conceptual phases of the project’s design,
magnitudes of probable cost should be assigned to this aspect of vulnerability reduction and a determination made regarding funding viability and
assignment. Inability to fund a competent protective structural design response
will inevitably lead to a commensurate increase in vulnerability and an increase
in the owner’s risk exposure.
68
DESIGN CONSIDERATIONS
Based on the project explosive event DBT values and the locations where
these threats can migrate within the project site and structure, a wide range of
structural protective design strategies may emerge. This may involve the introduction of additional structural members, increasing the size of structural elements, the introduction of new structural framing and load-carrying techniques
such as trusses and outrigger systems, addition and diversity of building materials, connection detailing, and other schemes, all of which may affect the overall appearance of the structural framing systems. Based on the aesthetic role
of structure in the particular project’s architectural vocabulary, there may be a
direct visual translation of these protective design influences on the architect’s
visual design intent. During the initial Conceptual phases of the project’s design, blast consultants and structural engineers should participate in design
charrettes with the project architects. These charrettes should address the
symbiotic relationship between the structural scheme options and architectural design intent to achieve the required blast protection and the desired
structural component of the architectural concept. Failure to address this
vulnerability reduction opportunity at this early phase within the project
may prejudice the opportunity to develop the most architecturally acceptable, security-competent, and cost-effective structural scheme.
While visual aesthetics often dominate design decisions, structural designs
responding to the requirement for blast mitigation may involve potentially significant impacts on programmatic space use, size, and arrangement. Increased
member sizes for blast protection, the addition of structural members or systems,
and/or the introduction of physically enforced standoff to improve structural survivability will, of necessity, reduce space available for other functional purposes.
Also, large structural systems such as trusses, additional shear walls, compression and tension rings, outrigger systems, and other similar systems may be of
such size and/or space invasiveness as to dictate relocation of architectural space
programs and/or MEP systems to floors other than where these structural systems need to occur to provide the required levels of protection. These space constraints may not be the only motivation to reassign architectural use programs or
MEP systems to alternate locations within the system. Based on the facility security program, the introduction of the threat and risk assessment identified DBTs
may place these critical structural systems at an enhanced level of risk. This occurs because it may not be possible to preclude (due to the space utilization and
occupant activity programs) the MPIED or VBIED from having access to the
desired locations where these structural system enhancements optimally would
occur. Consequently, during the initial Conceptual phases of the project’s
design, the blast consultant, structural engineers, and security professionals should participate in design charrettes with the project architects so that
there is agreement upon the ability to feasibly implement a security program
that precludes the explosive DBTs from having access to critical portions of
the structure. Also, these sessions should resolve the identification of structural impositions on MEP systems’ design and deployment, and the anticipated space impacts, both in terms of space availability and occupancy use
VULNERABILITY REDUCTION
69
programs. Failure to address these structural vulnerability reduction opportunities, at this early phase within the project, may prejudice the opportunity to develop the most competent and cost-effective structural schemes,
architectural designs, space utilizations, and MEP systems’ placement.
Value engineering has become an inherent project management process
throughout the design and construction profession. It has been shown to provide
beneficial results by optimizing the engineers’ design intent through a process
involving the application of previous experiences through a peer review by other
design professionals, contractors, and/or material fabricators. As long as this process continues, opportunities exist to modify structural systems in a manner that
may or may not reduce their ability to provide the enhanced protection they must
provide when they are used as part of an explosive event facility vulnerability
reduction strategy. The science of designing structural systems to resist explosive event loading is a unique and less well understood art than the management
of wind, gravity, and seismic loading. It is extremely likely that those involved
in the value engineering process may not fully understand the extent of blast
load structural system investment included by the original design engineers, and
recommendations for materials and/or structural systems’ design optimization
may be based upon inadequate knowledge or experience. Any value engineering and/or other similar “design optimization” process that involves review
and recommendations regarding structural systems engaged in blast vulnerability management should include the blast consultant and the structural
engineers responsible for designing the systems under evaluation.
Furthermore, during the construction phase of activity, the structural engineers
should remain involved in the shop drawing approval process to ensure that the
materials’ fabrication, erection, and field installation techniques and details continue to respond to the analytically proven designs that were developed as part
of the original blast mitigation work to achieve the required project vulnerability
reductions to the threat-assessment-identified explosive DBTs. Value engineering and quality assurance involvement during the construction phase by the
blast consultant is essential. As a result of the blast consultant’s analysis
and recommendations to protect against an explosive event, the structural
materials’ selection and the systems’ design and detailing may well be, of
necessity, more robust than that otherwise required for traditional gravity,
wind, and seismic loadings. Without the benefit of this blast analysis, the
rationale for unconventional material oversizing, encasement, overly robust
connections, and other structural enhancements may appear to be simply
overly conservative structural design and detailing, and may thereby create
an intuitive but false opportunity for invalidated optimization by the contracting community.
Structural designs providing explosive event protection are codified within
the drawings and specifications used for both procurement and directing contractor construction-related activities. Of necessity, details associated with member sizing, connections, reinforcing detailing, plating, and other blast protection techniques must appear in adequate detail to inform bidding, shop drawing
70
DESIGN CONSIDERATIONS
development, and field work. The structural engineering professionals should
develop, with the project architect and ownership team, the means to ensure
the nondisclosure of, and maintenance of the confidentiality of, structural
enhancements intended for the purpose of reducing vulnerability to the
project explosive DBTs. This process should be in place prior to the development of any documentation, whether Conceptual designs, blast analytics, or
more detailed documentation normally associated with the Schematic, Design Development, and Construction Document phases of design work. This
should include a series of published policies and procedures to manage electronic
and hardcopy design information.
2.8 RISK ACCEPTANCE
It is universally acknowledged within and outside of the design profession that
no facility can actually be designed, constructed, and operated in such a manner
as to be totally risk-free. There are simply too many variables and unknowns that
can create exposures and vulnerabilities that cannot be foreseen or completely
mitigated. The challenge is to reach a rational and justifiable position of balance
among available resources to be applied to facility design, construction, fit-out,
and subsequent operations, and to do so based on the best contemporary knowledge available at the time of design and on reasonable foreseeable future vision.
These resources include finances, design talent, design time and schedule, site
opportunities and constraints, materials availability, contracting skill, construction schedules, post-construction operational staffing, intellectual resources to
manage operations and create and enforce viable policies and procedures, and
last but not least, client risk tolerance and commitment.
This section discusses the implications of designing facilities to meet explosive threats effectively with less emphasis on resource implications and with
the primary goal of achieving lower levels of risk. This approach is generally
considered to be “designing to address the threat” or, more simply, “design to
threat.” The emphasis being that vulnerability reduction and threat mitigation
are the leading programmatic requirements for consequence management and
are intended to dominate the protective design decision process.
This section also discusses the implications of implementing designs that include mitigating strategies as they can be afforded by available resources, and
the consequent potential increases in risk exposure if these resources are limited. This approach is generally considered to be “designing within budget” or,
more simply, “design to budget.” This remains as one of the more significant
challenges to the design professional and client team, as there are almost always limited resources available, and this then requires an eventual trade-off
between resource allocation and protective design implementation. Just as it
is unreasonable to assume that no design and facility operation will be devoid
of threats, vulnerabilities, and risks, so is it unrealistic to assume that there are
RISK ACCEPTANCE
71
unlimited resources available to achieve complete risk elimination. The process
of adjudicating this tension between resource consumption and facility security enhancements will require the design professionals to clearly identify
optional levels of protective enhancement so that client risk management
and the risk tolerance allocation process have a basis for rational decision
making.
2.8.1 Design to Threat
This subsection focuses on the process of assigning and allocating resource dedications adequate to fully address the threats identified in the threat and risk
assessment process as referenced in Section 2.4 and the associated advantages in
terms of increased opportunities for substantial risk reduction. Consequently, the
success of the design team in achieving the extent of risk acceptance identified
by ownership will, in part, be based upon a clear understanding of the projectspecific threats identified in the threat and risk assessment. As mentioned previously, for explosive events, this will require a clear and concise definition of the
details associated with the types of events. This may include purposeful criminal, man-made devices or naturally occurring opportunities for there to be an
explosion that places the facility at risk. For the criminal act, this will include the
type of explosive to be used for analytical modeling (usually, this is expressed
in terms of TNT equivalent), the size of the design basis threats (usually, this is
expressed in terms of a specific weight of TNT equivalent), the delivery mechanisms (MPIED, VBIED, etc.), and other specific characteristics (such as charge
shaping, etc.). It is important to note that a project may well have multiple DBTs
for a variety of different threat scenarios. For other explosive event opportunities,
well-defined criteria need to be established, including the source of the explosive agent, its containment or delivery vessel or system, quantities, and other
relevant specific data necessary to achieve the appropriate level of analytical
modeling.
Once the threat data are defined and the post-explosive-event facility performance criteria are fully understood and agreed to with ownership, the design
to threat approach requires the design professionals to engage in the process of
clearly identifying facility vulnerabilities. This will, of necessity, involve the services of security professionals so that the potential delivery of threats can be
clearly defined based on the architectural design, planned electronic and operational security countermeasures, and other explosive threat-reduction program
elements. This will allow a clear understanding of remaining areas of explosive
vulnerability to both criminal and noncriminal explosive threats. Since the resource investments required to substantially reduce risk from explosive events
are normally substantial, addressing the specific areas of greatest vulnerability and then identifying remaining areas of lesser vulnerability is prudent. This
process of prioritizing vulnerability from most significant to least significant
is positively dependent upon the design professionals clearly identifying, and
72
DESIGN CONSIDERATIONS
agreeing with ownership on, where an explosive threat can occur. The development of a rational and known to be competent security program, which
identifies with confidence limited areas of threat potential, is the most economical means to assign resources to limited areas of vulnerability mitigation. This will achieve the best risk reduction with the least resource
allocation.
While reducing the quantity of vulnerabilities has substantial virtues, those remaining vulnerabilities may be significant and challenging. The design to threat
approach, quite simply, requires the design professional to develop the appropriate mitigating strategies during the design process, so that the post-event performance criteria meet the standards originally established during the Conceptual
Design phase, as codified in the project key design assumptions. In anticipation
that required resource allocations will not be trivial, the design professional team
may find it most beneficial to develop two or three alternate schemes of structural system design, so that there is an inherent and preestablished response to
the formal or informal value engineering process, which will inevitably occur as
part of any well-managed project.
For the structural engineer, the design to threat process must assume that
threat reduction opportunities have been fully exploited. Simply stated, opportunities for reducing the DBTs have been taken into consideration by law enforcement agencies and by other project mitigation strategies developed by ownership and the balance of the design professionals. As an example, this may
include the provision of security personnel, equipment, and policies and procedures, such as the use of explosive detection devices and screening operations as
a means to reach an agreed-to explosive DBT level at various points throughout
the site and facility. Assumptions involving the definitive reduction of explosive DBTs and their potential delivery locations should be specifically validated by a well-orchestrated and documented series of design charrettes
involving ownership, the architect, the security professional, and the structural engineer.
The design to threat approach may, in fact, inform ownership that structural system protection, hardening, or alternate load transfer path designs create substantial resource allocation hardships. Irrespective of the development
and presentation of various structural design options intended to address
the explosive threat, it may be necessary to carefully examine the means
to reduce the threat, reduce the number of vulnerabilities, or have ownership participate in greater risk acceptance. This may occur as a result of the
unattractive and/or unmanageable financial or programmatic implications of the
structural protective design options. It is for this reason that addressing the opportunity to employ the design to threat approach during the earliest Conceptual
and Schematic design phases is essential. The viability of designing to specifically achieve the required threat reduction through a process of structural design
mitigation may identify the need to impose and implement the other threat mitigation resources. This may include the development of protective architectural
and site civil design features, deployment of security systems, equipment, and
RISK ACCEPTANCE
73
manpower, and recurring capital expenditures for operations. It may also involve
increased restrictions associated with space-use programs and/or redesign of the
systems associated with mitigating the effects of accidental explosive events.
2.8.2 Design to Budget
This subsection discusses the counterpoint to the design to threat process. Specifically, the design to budget process can be invoked by ownership when they
determine that finite resources are available and a set quantity is established;
the extent of protection provided is then tailored to achieve the best risk reduction benefit within those strictly defined resource boundaries. The term “budget”
in this section of the handbook refers to the budget of overall resources, a definition purposefully intended to be more than just project funding. However, more
often than not, the project resources that will have the most confining effect on
protection strategies intended to mitigate the explosive event will be the project
funds available to spend on security. Occasionally, other resource issues, such
as lack of specific protective design materials and/or the skilled labor to install
them, will also create resource constraints and may direct the design professional to utilize an overabundance of those protective strategies that are the most
available. As an example, the inability to include a competent electronic security
system as part of the project, whether due to lack of skilled labor, system components, or the ability for that system to be reliably maintained and serviced, may
dictate that explosive event facility vulnerabilities require greater investments
in physical hardening to make up for the inability to implement early warning
detection and control of access.
Once there is agreement on the project definition of budget resources and the
extent of funding available for an agreed-to definition of “project security,” the
design professionals should conduct a series of design charrettes. At a minimum,
these should include the project architect, structural engineer, blast consultant,
MEP and vertical transportation consultants, site civil and landscape designers, and security professionals. Other consultants, including traffic engineers,
lighting consultants, geotechnical engineers, and others may need to participate
as well, based on the project’s location, occupancy program, and complexity.
The intent in sequestering the design teams’ resources is to create a series
of consensus-developed options (at least two but preferably three) wherein
each design discipline inherits a small portion of the responsibility for mitigating the facility’s vulnerability to explosives. This has repeatedly been
shown to be the least expensive and most effective approach to resource allocation and will provide the best opportunity to meet both protective design
and budget objectives. In short, this is the most expeditious path toward “designing to budget.”
Unfortunately, these design charrettes, while yielding the most comprehensive
and effective solutions at the least cost, often still identify funding expenditures
in excess of client resources. This is usually a product of design professional conservatism, wherein each professional advances protective designs that have the
74
DESIGN CONSIDERATIONS
absolute least opportunity to fail at that professional’s particular mission. Ownership’s misallocation of funding (underfunding), due to their extremely limited
experience and understanding of the investments required to create truly credible explosive event consequence reduction designs, is also a factor. Sometimes,
however, there are simply insufficient overall project funds to create great architecture that includes spaces that are intended to support continued critical
operations in environments of considerable threat potential.
When there is a disparity between the ownership’s available resource pool
and the resource requirement dedications to meet the originally agreed-to postexplosive-event performance, the design to budget process approach does not
need to be abandoned. However, the design approaches that remain will all involve an extent of risk acceptance by ownership. Quite simply, at this point, there
is an admission by the design and ownership teams that vulnerabilities will continue to exist in the project that may not be adequately addressed. In order to
achieve the best result, the project vulnerabilities to the explosive event DBTs
identified in the threat and risk assessment will need to be listed and prioritized
based on the anticipated consequential effects of an explosive event. This can
be most successfully accomplished if there is a comprehensive and prioritized
list of consequential effects. A thorough understanding of the prioritized vulnerabilities and prioritized consequential effects will allow the available project
resources to be applied in a manner that provides the best extent of protection
and risk management. Once a determination has been made about how limited project resources are to be allocated, a concise written protective design
narrative should be authored by the design team. This narrative should be
subsequently reviewed and approved by ownership so that there is a legacy
document that clearly explains the rationale behind resource allocation and
the extent of explosive event protection provided.
It is essential for the design professionals to understand that client bracketing
of funds for project security will need to cover investments other than facade,
structural systems, ERR systems, and sensitive space hardening. Security strategies to reduce the explosive event threat will also require expenditures for site
civil vehicle approach interdiction features, electronic security systems, both personnel and vehicular screening equipment, and the costs for security staffing and
the build-out of their programmatic support spaces. Based on the project costestimating techniques and/or models used, redundant diverse ERR systems and
architectural features, and redundant diverse business continuity systems may be
lumped into the overall security budget. In order for the design to budget process to be effective—and in order to avoid mismanagement of the extent of
funding available for actual explosive event security mitigating building features, components, and systems—the design professionals must engage in a
reconciliation process with ownership. This partnership should address and
agree upon what project features and components constitute an explosive
event protective design strategy and the budget available to achieve it. This
activity should take place at the time when the project budget is in the process of
being established.
RECENT EXAMPLES OF SECURITY DESIGN “BEST PRACTICES”
75
2.9 SOME RECENT EXAMPLES OF SECURITY DESIGN “BEST
PRACTICES”
The authors’ combined 80-plus years of security experience have acquainted
them with a broad range of effective and workable protective building-design
concepts and mitigation solutions, summarized below, that are in daily use
around the world. (Because of the sensitivities surrounding post-9/11 security,
venues are not specified.8 ) Allocating space for these measures, and incorporating utility requirements for them, are considerations that security design teams
need to address at the earliest design stages.
1. Defense in Depth—Creating a layered zone defense built upon concentric
rings of interlocking physical security, security systems and equipment,
personnel, and policies and procedures.
2. Security Zones of Control—Enforcing a distinction between different physical spaces and the level of security assurance that each space must provide.
Some locations are designated as sterile; that is, everyone in them has been
screened for explosives (as well as chemical/biological/radiological materials and other hazardous or prohibited items). Other spaces are identified
as secure; that is, only people with proper access privileges are permitted
to enter. Other areas may be open to the general public and without restrictions. Additional zones of control may also exist based on the project type,
size, and location.
3. Redundancy and Diversity—Designing for the redundancy and diversity
of critical assets, including security, ERR, and MEP systems, as well as
client-specific physical and intellectual assets.
4. Exceptionally Protected Zones—Specifically designing for the preservation of unique cultural artifacts and the protection of unique financial assets, intellectual resources, critical unique industrial or financial processes,
and other single-source assets that do not allow for protection through diverse repetition. To protect one-of-a-kind assets, hardening solutions have
to be created in a manner responsive specifically to each asset’s vulnerability.
5. Personnel Screening—Identifying and screening all individuals and their
personal items for explosives and other proscribed or hazardous materials.
8
These sites include the U.S. Capitol, the Pentagon, Central Intelligence Agency (CIA) Headquarters, Boston’s Logan International Airport, the London Ring of Steel, London’s Canary Wharf commercial office complex, and the Eurotunnel between the United Kingdom and France. This best
practices section also includes observations from the Sears Tower in Chicago and the Statue of
Liberty Visitor Screening Center. Relevant security measures at a worldwide insurance company
headquartered in lower Manhattan and a number of comparable private sector enterprises, as well
as several U.S. border inspection stations, are also incorporated.
76
DESIGN CONSIDERATIONS
6. Vehicle Screening—Identifying and screening all vehicles to prevent
vehicle-borne explosive (and other) threats from entering managed street
segments, on-site parking, subgrade parking, and cargo offloading areas.
7. Cargo and Mail Screening—Ensuring that qualified and trained personnel
screen for explosives all cargo deliveries destined for the buildings’ core
interior areas in accordance with the facility-specific DBT. For example,
security personnel would inspect all overnight and freight couriers, and
other nonpostal/mail services for explosives and other threats. In some very
high-security venues, all delivery vehicles, mail, and cargo may need to be
inspected off-site before delivery to the protected premises.
8. Secondary Screening Capability—Enforcing a clearly defined separation
between the primary inspection process for screening people and vehicles
and detecting anomalies, and the secondary inspection process, which resolves any anomalies through detailed search, investigation, and testing.
At least two different detection technologies should be used to complement each other—for example, using explosive-detection canines, explosive vapor sampling, trace detection, or detailed visual inspections to conduct secondary investigations when high-energy imaging appears to detect
explosive materials.
9. Space and Provisioning for Security Control Centers—Ensuring that the
facility command, control, communications, and intelligence (C3I) center
that monitors and controls security systems and operations is afforded the
appropriate space, utilities, security, environmental protection (from smoke
and chemical/biological agents, as appropriate), and backup systems.
10. Segregation and Protection of Building Mechanical, Security, and ERR
Equipment—Isolating critical security, ERR, and MEP systems and areas
from tampering and disabling by unauthorized individuals.
11. Crime Prevention through Environmental Design (CPTED)—Designing
CPTED—a concept that incorporates environmental design features
into the overall security strategy for deterring crime and unwanted
behavior—into an area or facility.
2.10 RELATED PHENOMENA
Designing to comprehensively mitigate the explosive event requires an understanding of the consequential effects associated with the release of significant
amounts of energy over very short durations. Design professionals must be cognizant that the explosive event alone is not the only threat consequence requiring
design attention and subsequent mitigation. Post-explosive-event effects may be
numerous based on the type of explosive event, location internal or external to
the facility, the type of building construction, internal contents, occupancy use
program, and numerous other contributing factors such as the flame, smoke, and
fuel contribution of the structure itself and its contents.
RELATED PHENOMENA
77
No discussion of “related phenomena” would be complete without the acknowledgment that there will be the potential for severe injuries and fatalities;
damage to critical facility assets, both intellectual and physical; disruption of
business missions; and the need for implementing evacuation, rescue, and recovery operations. Additional effects will include crime scene management,
post-event facility evaluations, insurance claim investigations, and reconstruction
and restoration of physical plant, institutional operational viability, and public
image.
However, design professionals have the opportunity to influence a limited
venue of related explosive event phenomena. With appropriate assistance from
blast and security consultants, engineers can design to contain and manage fires
which may follow some explosions. Also, limitations to the extent of post-event
damage and structural system collapse can be analyzed and responded to, so that
the structural performance of the facility is engaged in a limited participation
in life-safety, ERR, and business continuity exposures. Well-developed facade
designs can reduce injuries and fatalities while also protecting interior spaces,
the ERR, and critical business continuity systems. Design professionals can also
identify the intended means of evacuating building populations and engage in
the planning and providing for the subsequent arrival of emergency responders
so that effective ERR can enhance the opportunity for occupant safety. In doing
so, they can clearly identify those building features, systems, and components
that require an enhanced level of explosive event protection as it is extremely
likely that they will be severely compromised by such an event unless they receive extraordinary protection.
Of the numerous explosive-event-related phenomena, the following handbook
sections emphasize the need for the design team to pay particular attention to
and dedicate both design and building resources to the mitigation of progressive
collapse, protection of ERR, and management of attendant fires.
2.10.1 Progressive Collapse
As mentioned earlier in this chapter of the handbook, the losses of primary and
secondary components of the building’s structure are potential results of an explosive event, whether a purposeful and malicious act or an accidental or naturally occurring incident. The extent of damage, as previously noted, will depend
upon the size and type of event, its location, and the extent to which the structural
system in the building has been prepared for managing the extraordinary applied
loads and any significant impacts due to airborne debris.
Since life safety is generally the paramount criterion, the definition of disproportionate damage used to define the project’s post-explosive-event security performance criteria may be dependent upon the size and location of the
building’s occupant population. High-rise office buildings with commercial occupancies traditionally allocate 150 to 200 square feet per person as a means
to estimate population densities. With common floor plate sizes ranging from
25,000 to 45,000 square feet, there are opportunities for approximately 150 to
78
DESIGN CONSIDERATIONS
300 people per floor. The loss of multiple floors, especially throughout the entire
height of the building, which may be as many as 70 or 80 floors, defines catastrophic loss of life circumstances. Hence, progressive collapse from one floor
to the next is deemed unacceptable. However, the definition of disproportionate damage may be more conservative than the prevention of damage outside
of the area of explicit and direct explosive event effects. Design professionals
should, at the earliest possible design phase in the project, review with the
architect the anticipated occupancy profiles for the building’s interior and
exterior spaces. They should examine the extent of failure associated with
the threat-assessment-identified explosive event DBTs or other possible explosive sources within the facility, should no supplemental structural robustness be added for explosive event management. At this point, a determination should be made whether the consequent resulting structural damage
invokes too high a casualty count for the owner’s and design professionals’
risk tolerance. If that is in fact the case, then while the damage may not be
structurally disproportionate, the life-safety consequences may be unacceptable.
This may then inform a more conservative definition of allowable disproportionate damage. Since the design professionals, from a structural standpoint,
generally are responsible for designing to preclude progressive collapse and
to achieve nondisproportionate damage, clear definitions of these terms, especially disproportionate damage, should be reviewed with ownership and
codified as part of the project’s key design assumptions. The significance of
these definitions becomes even more evident for low-rise facilities that have significant occupancies concentrated in relatively small areas, such as performance
venues or transportation structures. In these environments, there may be, for estimating purposes, one person for every 4 to 6 square feet, which may translate into thousands of persons within a relatively small area. Again, a traditional
structural definition of disproportionate damage may define acceptable structural
damage to four column bays as a result of the loss of a single column. However,
the loss of these bays may translate to, using a 40-foot-square bay module, 6,400
square feet of occupancy. In this example, at a density of 6 square feet per person, there are potential casualties of over 1,000 people. This is an unacceptable
result. Therefore, each project must be individually evaluated and the progressive
collapse designs developed accordingly.
In order to competently evaluate either preexisting structures intended for
renovation and/or expansion or the design of new structures, the services of a
qualified blast consultant working in partnership with the structural engineer are
essential. The determination of the extent of anticipated damage and the opportunities to identify the extent of potential mitigation will be based upon their
interactive use of structural analysis, blast effects science and software, historically informed engineering judgment, and the validation of results conclusively
drawn from these experiences and analytical tools through actual live testing of
structural systems with actual explosive events. Failure to properly anticipate
and define the circumstances and designs that will not prevent progressive
RELATED PHENOMENA
79
collapse and that will not prevent disproportionate damage will preclude the
elimination of potentially fatal flaws in the design and place both ownership
and the design professionals at considerable risk.
2.10.2 Disruption of Evacuation, Rescue, and Recovery Systems
Careful study of post-explosive-event conditions clearly identify the value provided by building designs that provide a safe and secure means of egress for
internal occupants and equally safe and secure means of reentry for emergency
responders. Time-motion egress analytics, which model the amount of time
required for a building to engage in a complete evacuation of all its occupants, routinely show dramatic improvements when all of the intended
means of egress are available and unimpaired. Or, stated another way, the
loss of egress capacity associated with an explosive event can be expected to
dramatically increase the time required for people to exit a facility at risk and to
substantially diminish emergency responder efficacy at extracting the base population of handicapped individuals and those incapacitated by the explosive event
itself. This can be expected to translate into higher injury and casualty counts, as
every moment is crucial during facility occupant extraction.
Structural integrity of stairs and elevators (and sometimes escalators, as these
may also be used as an egress mechanism in some facilities) may not translate
to a tenable ERR environment. As exemplified by the 1993 use of explosives at
the World Trade Center by terrorists, stairwells became filled with smoke, even
though they were structurally adequate to continue to function as viable means
of egress. This smoke condition hindered exiting operations and was allegedly
responsible for smoke inhalation, and a significant number of related adverse
health consequences were claimed. Structural integrity maybe inadequate if not
supported by survivable smoke control/pressurization systems combined with
interlocked and functionally survivable smoke detection.
At the time of authorship of this handbook, there is a distinct and evolving
trend for building codes and other legislative documents that guide and/or direct
the design community to require the survivability of evacuation, rescue, and recovery systems after an explosive event. Considerable attention has been focused
on this subject as a result of the extensive studies carried out by the National Institute of Standards and Technology (NIST) of the World Trade Center attacks
of September 11, 2001. Numerous building safety recommendations were made
by the NIST committee, many of which focus on a need for additional legislation and design responsiveness to the enhancement and preservation of ERR
systems, building features, and design components. It is reasonable to expect,
and this handbook anticipates, that these NIST recommendations will become
the foundation for new language in national and international building codes. It
is reasonable to assume that such significance has been focused on ERR survivability that these codes will call for increased stairwell widths; more robust
stairwell construction; the use of elevators for ERR operations, and consequently
80
DESIGN CONSIDERATIONS
the construction of more robust and water-resistant hoistways and lifts; enhanced
fireproofing ratings, adhesiveness and impact resistance; improved fire detection
and reporting; more survivable and more competent fire suppression systems;
more effective emergency voice communication systems; and other related support infrastructure improvements. With these known to be valuable and soon
to be codified ERR building features, design professionals should acknowledge that one of the most significant explosive-event-related phenomena requiring attention is the preservation of facility ERR operations.
In Section 2.5.2, a comprehensive list of ERR systems was provided. While
it is not reasonable to expect that these systems will survive in the immediate
zone of the most powerful explosive event effects, design professionals should
carefully analyze the location where the threat-assessment-identified explosive
event DBTs can occur. They should then design ERR systems so that they are
appropriately protected to achieve survivability and continue to provide functionality to the balance of the facility. This may involve protective hardening
of the spaces where they occur or the design of redundant systems with adequate geographical diversity so that they are not disabled by the DBT events.
To accomplish this effectively is no small design endeavor. A quick review of
the ERR building components will intuitively identify that they are likely to be
spread throughout the entire facility. This may therefore require a detailed review of the entire facility and all of the attendant systems and analytical modeling of the protective design features in response to all of the various explosive
DBTs. Consequently, the design professionals should carefully review the
client’s expectations, code requirements, and good engineering practice and
develop a clearly stated standard with respect to the desired performance
of ERR systems and building features in response to the project’s explosive DBTs. To do this competently, it must be remembered that many systems
that support ERR are functionally dependent upon each other. As an example,
stair pressurization, smoke purge, and smoke control fans, necessary to maintain egress and firefighting, are integrally linked to the building management
system (BMS), fire command systems, and emergency power, and the compromise of any one of those by the explosive event will render that system inoperative. Consequently, all of these systems must be systematically analyzed and
protected. As another example, fire suppression systems, necessary to manage
potential attendant fires, are dependent upon the integrity of fire suppression
piping, pumps and control panels, emergency power, and fire alarm systems.
Again, a compromise of one of these components may place the entire system
at risk.
There are simply too many permutations of ERR systems integration that
could be placed at risk by explosive events to competently note and review each
of them individually in this handbook for the purpose of developing uniquely responsive mitigating strategies. The appropriate process to ensure that ERR systems have the greatest probability of post-event survivability is further discussed
elsewhere in Chapter 2. as part of consequence management, and is founded
upon the strategy of overlaying ERR systems on the project threat maps and
RELATED PHENOMENA
81
having their designs and/or hardening protection informed by individual explosive event modeling using the project DBTs.
2.10.3 Attendant Fires
Not all explosive events are accompanied by an immediately ensuing fire development. The rapid expansion of gases at hypersonic speeds and the instantaneous
consumption of oxygen as part of the explosive event chemistry often deny fire
development. To that end, the oil industry has used oxygen starvation as a means
of suppressing oil well fires through the use of explosives. However, certain types
of explosive events, such as those caused by high-pressure gas mains, often have
an initial burst of high energy but then continue to burn at extreme temperatures and have the opportunity to subsequently engage combustible materials or
continue to generate adequate heat from their own fuel source to subject structural elements to ductile yielding and reduction in load-carrying capacity. Mitigating strategies often involve the opportunity to eliminate the participation of
secondary fuel sources ignited by an explosive event. The design professionals
should carefully review the storage and distribution systems for flammable
agents and their exposures to the project explosive DBTs and include in
their design pump shutdowns, automatic and manual breach control valves,
and/or other strategies so that these fuel sources have limited burn times.
Additionally, the design of robust encasement around fuel storage and/or
distribution systems should be considered. Examples include the installation
of generator fuel storage reserves in what are essentially concrete bunkers providing both explosive event ignition protection and elimination of secondarysource fuel supply.
Attendant fires pose a range of problems. As mentioned previously, they may
challenge the structural integrity of the facility as a result of extreme temperatures, especially applied in an environment where fireproofing may be damaged
and/or absent as a result of the initial blast pressures and abrasion/impact of airborne debris associated with the explosive event. Although it is anticipated that
building code legislation will likely require increased fire ratings for structural
members and the specification of more physically resilient fireproofing methods, it is unlikely that these material specifications will provide adequate postexplosive-event performance reliability. The structural engineer and architect
should partner with the project blast consultant and security professional
and determine the opportunity to receive enhanced benefits from the specification of robust fireproofing strategies. This may include the use of concrete or
masonry encasement. Reasonable solutions for structural system protection from
fire will continue to include fireproofing, fire suppression systems, and alternate
load path designs. However, the most effective mitigating strategy remains
the reduction and/or elimination of the explosive threat, and the reduction
of building materials and contents that have the opportunity to participate
as fuel sources in the fire development triad of oxygen, fuel load, and an
ignition source.
82
DESIGN CONSIDERATIONS
The National Fire Protection AssociationStandards has quantified the parameters by which building occupants are affected by fire events. Temperature, air
toxicity levels, and visual obscurescence are all factors that affect either health
or the ability to extricate oneself from the fire event. Explosive events that are
accompanied by attendant fires are likely to be characterized by significant quantities of airborne debris, irrespective of the smoke contribution associated with
the fire. Air quality may rapidly approach lethal thresholds as a result of escalating air temperatures (140◦ F is the upper-level threshold for human exposure). Air
quality may also become untenable as a result of toxic air particle concentration
from the explosive demolition effects and/or from the fire’s oxygen consumption
or toxic gas generation. These conditions combined with visual obscurescence
create a potentially deadly environment. This situation and the previous concern
regarding structural integrity are exacerbated by the likely failure of sprinkler
or gaseous suppression systems in the immediate area of the event. However,
containing the event-based fire to a smaller radius of effect will do much
to improve life safety for those outside of the explosive event zone. Consequently, the design of sprinkler systems with smaller fire control zones,
increased water application density, fast response sprinkler heads, and survivable water supplies provides a very reasonable and competent means to
both improve life safety and reduce property damage. As part of any explosive event protection strategy, the blast consultant and the structural engineer
should partner with the fire suppression engineer for the purpose of optimizing
the sprinkler system’s response to an explosive event. This should include sprinkler coverage zone sizing; the use of automatic and/or manual breach control
valves, in conjunction with alternate main and/or pipe riser cross connections,
to ensure system performance outside of the explosive event zone; increased
hazard classifications to improve water availability and discharge density; head
selection; and other pertinent aspects of suppression systems’ design with the
intention of fire containment to a limited area. This will have the dual benefit of
reducing loss of life and property and reducing the extent of structure exposed to
extreme heat.
There may be less opportunity to influence the selection of the actual building
construction materials and the interior fit-out of furniture, fixtures, and equipment, to reduce the building’s internal fuel load, but, as an explosive event design
mitigation strategy, it should not be summarily overlooked. The less there is to
burn, the shorter the duration and overall exposure of the building’s occupants
and structural elements to the deleterious effects of a fire born out of an explosive
event. Similar design logic follows in the selection of building construction materials. The specification of noncombustible components will reduce fire damage.
Generally, structural systems and building contents are regulated by the Building
and Fire Codes with the intent to reduce combustibility. However, whenever the
project threat assessment identifies explosive events as a threat to be mitigated, building and interior fit-out components should be specified with an
eye toward reducing the facility fuel load.
SECURITY DESIGN CONSIDERATION GUIDELINES
83
2.11 SECURITY DESIGN CONSIDERATION GUIDELINES
The following bullets summarize major points in the building security design
considerations process in a condensed format.
1. During the project’s design’s initial Conceptual phases, the blast consultant and structural engineers should participate in design charrettes with
the project architects so that the symbiotic relationship between the structural scheme options and the architectural design concepts to achieve the
required explosive event protective designs are considered for potential implementation. Failure to address this vulnerability reduction opportunity
at this early phase may prejudice the opportunity to develop the most architecturally acceptable, security-competent, and cost-effective structural
scheme.
2. The entire security design team should participate in the creation of a brief
design mission statement, which clearly and concisely states the client’s
expectations for the facility and site’s post-explosive-event performance.
Since these expectations are routinely overstated by clients—as a result of
their lack of sophistication or knowledge of the engineering complexities,
architectural program compromises, post-construction institutional operational constraints, impact on project financials, aesthetic effects, and other
factors better understood by the structural engineer and the balance of the
design team professionals—this mission statement must be a consensusdeveloped document.
3. On projects where explosive event consequence management is a programmatic design response requirement, design professionals should insist that
the threat assessment and site-specific DBT be provided to them at the earliest possible date. This is essential if the specific site and facility vulnerabilities are to be identified and mitigating concepts developed for potential
implementation.
4. During the site selection process, the security and blast consultant explosive event engineering professionals should work closely with the
client as full members of the building security design team. Based on
the maximum realistic standoffs, the pressures and impulses that will
be applied to the structure, its facade, inbound utilities, and other critical site civil features should be factored into the site selection decision. This is essential so that an initial determination can be made on
how these extraordinary loads (in excess of gravity, wind, and seismic) will affect the protective design capabilities of the major building
systems.
5. The initial phases of consequence evaluation and management should occur in the Conceptual and Schematic Design phases and include a review
84
DESIGN CONSIDERATIONS
6.
7.
8.
9.
10.
of the opportunities for redundancy and diversity and the use of hardened
facility structural, facade, and ERR design elements.
Insufficient efforts to clearly define the client’s consequence management
expectations are one of the most significant design process shortfalls and
are a major reason why the engineering efforts fail to meet client needs or
expectations.
Deliberate or accidental explosive events rarely impact only the areas defined within contract boundary limits of work, a fact that must be taken into
consideration as part of consequence evaluations. During the initial stages
of site planning and facility design blocking, stacking, and facade selection, consequence evaluations should consider the aggressors’ opportunity
to deliver the explosive to the target and the extent of damage a detonation
external to the building or site could cause.
Design professionals must address building designs so that the explosive
event consequence management response is based on an evaluation of how
ERR will occur. In this way, designs are informed to protect ERR building
design features and provide inherent tactical support and definitive criteria
for an ERR concept of operations.
While the pressures, impulses, and plasma effects associated with an explosive event using extremely high-energy substances such as plastic explosives are regarded as upper-level protection design thresholds, the security
design team should carefully review the entire project design to identify
other building systems or storage spaces that could participate in an explosive event.
Any value engineering or other similar “design optimization” process
that involves review and recommendations regarding structural systems
engaged in blast vulnerability management should include the blast
consultant and the structural engineers responsible for the systems under
evaluation.
2.12 CONCLUSION
It is the authors’ hope that this discussion of security design considerations has
given the reader an increased awareness of how a competent and well-organized
security design approach, initiated at the beginning of the design process, can
significantly influence the eventual success or failure of the facility’s explosive
event protective design strategies. Managing the design and security interdependencies and complexities of today’s blast-resistant construction, in an uncertain
and ever changing threat environment, requires the contributions of security, explosives, and building sciences domain experts working together with ownership
risk managers, forming a group that can then appropriately be called the security
design team.
REFERENCES
85
REFERENCES
Boettcher, Mike and Ingrid Arnesen. 2002. Al-Qaeda Documents Outline Serious
Weapons Program. CNN.com (January 25, 2002), available at http://www.isisonline.org/publications/terrorism/cnnstory.html.
Federal Emergency Management Agency. 2005. Risk Assessment: A How-To Guide
to Mitigate Potential Terrorist Attacks (FEMA 452). Washington, DC: Federal
Emergency Management Agency, Department of Homeland Security. Available at
http://www.fema.gov/plan/prevent/rms/rmsp452.shtm.
Robertson, Nic. 2002. Bomb-making video reveals scope of al Qaeda threat. CNN.com
(August 21, 2002), available at http://archives.cnn.com/2002/US/08/21/terror.tape.
main/index.html.
Santayana, George. 1905. The Life of Reason, or the Phases of Human Progress: Introduction and Reason in Common Sense. New York: Charles Scribner’s Sons.
U.S. Department of Defense. DOD Dictionary of Military Terms. http://www.dtic.
mil/doctrine/jel/doddict/.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
3
Performance Criteria for
Blast-Resistant Structural
Components
Charles J. Oswald
3.1 INTRODUCTION
Performance criteria specify quantitative limits on the response of structural
components that are intended to achieve the blast design objective or strategy for
the building design, such as those summarized in Table 3.1. A structural component is designed by calculating the blast load on the component, using a dynamic
analysis to determine the maximum component response, and then checking the
response against quantitative limits that are consistent with the overall building
design objective.
In typical static design, the stress level of component response is limited to
prevent failure. However, this approach cannot be used for blast design, since
controlled, ductile yielding is usually part of the design intent. Therefore, response limitations are typically placed on the maximum dynamic deflection of
the component to prevent component failure, where this deflection accounts in
an approximate manner for the amount of acceptable damage or plastic strain.
The amount of conservatism in the design depends on the allowable component deflection limits. Usually, the allowable component deflection is limited so
that either repairable, or unrepairable, damage occurs, but not component failure.
In other cases, when component failure from a worst-case explosion scenario is
almost impossible to prevent, response limitations may be placed on the postfailure velocity of the components and the amount of overall building collapse.
Table 3.2 shows different areas of blast design and typical performance goals
for each area. The performance goals are affected primarily by the type of blast
threat for each area. For example, terrorist attacks involve many uncertainties related to the explosive threat, and consequently this area of blast design has more
basic design performance goals. Explosive safety, on the other hand, involves explosions of known stored or manufactured materials, and its performance goals
have more defined levels of protection. Hardened military structures typically
contain mission-critical equipment and personnel (i.e., command and control
87
88
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.1
Summary of Blast Design Strategies
Performance Goals
Design Strategy
Comments
Prevent component
failure
Limit response of components
protecting personnel and
equipment to levels below those
causing failure. A safety factor
against failure is provided by
minimizing expected damage
levels.
This is the primary approach used
to protect building occupants
against injuries from an
explosion.
Limit structural
collapse
Prescribe more conservative
allowable response criteria for
load-bearing components. Also,
design buildings to withstand
localized failure of a primary or
load-bearing component (i.e.,
design against progressive
collapse).
UFC 4-023-03 (UFC 2005) and
GSA (GSA 2003) have specific
criteria to design against
progressive collapse. Building
collapse will almost always
cause a very high percentage of
fatalities to building occupants.
Maintain building
envelope
Design against failure of building
cladding elements, including
windows and doors.
Blast overpressure entering
buildings can cause significant
injuries and extensive damage,
but not usually many fatalities.
Minimize flying
debris
Design cladding components using an Mitigating the extent and severity
of injuries to occupants in
expected worst-case explosive
building areas without overall
threat to fail without becoming
structure collapse is very
hazardous projectiles to the
dependent on minimizing flying
occupied space. This may be used
debris.
for a limited building area receiving
the worst-case blast loads.
Prevent cascading
explosion events
Design structures around stored or
manufactured explosives to resist
projectiles and blast pressures from
a nearby explosion without failing
catastrophically and detonating
explosives within structure. Also,
design building components to
protect personnel and control
equipment necessary for shutting
down potentially explosive
surrounding industrial processes in
an explosion.
This includes proscriptive
requirements based on testing in
DDESB 6055.9 (DDESB 2004)
and blast design requirements to
protect equipment in UFC
3-340-02 (UFC 2008) for cases
involving high explosives and
ASCE (ASCE 1997b) for cases
involving industrial explosions.
centers) that must be protected against a wide array of enemy threats, but these
threats are relatively well defined.
3.2 BUILDING AND COMPONENT PERFORMANCE CRITERIA
Blast design and analysis are primarily component-based, whereby the applied
blast load and response of each component in the building are determined
BUILDING AND COMPONENT PERFORMANCE CRITERIA
89
Table 3.2 Summary of Blast Design Areas
Design Area
Performance Goals
Comments
Explosive safety
Protect personnel
Protect equipment,
supplies, and stored
explosives Prevent
cascading explosive
events
Design against accidental explosion where
the amount and location of potential
explosives are relatively well quantified.
All inhabited areas are designed to
prevent injuries. This type of blast design
generally has the most design
conservatism.
Hardened military
structures
Preserve
mission-critical
functions
Buildings designed to resist attacks from
specified weapons with a near-miss or
direct hit such that given operations can
still be performed.
Antiterrorism
Prevent mass casualties
Minimize flying
debris Limit
structural collapse
There are uncertainties in the amount and
location of potential explosions caused
by terrorists. The usual intent is to build
in a given amount of blast resistance
using a level of protection against a given
defined threat. A maximum standoff
distance to the protected building is
established with barriers and inspections
for explosives.
Process safety
Protect personnel
Prevent cascading
events Minimize
financial losses
Explosive events can be modeled relatively
well, but initial conditions of the accident
causing explosion must be assumed.
Neutral risk philosophy is employed such
that people inside buildings should have
the same protection as those outside
buildings. Ability to shut down
operations is often important to prevent
cascading events.
individually. However, performance goals are usually set in terms of life safety,
functionality, and reusability for the entire building. Therefore, damage or response levels for individual building components must be established to achieve
the overall building performance goal. A framework for this process has been developed by the American Society of Civil Engineers (ASCE, forthcoming) based
on a similar approach developed by the U.S. Army Corps of Engineers, Protective Design Center (PDC) (Protective Design Center 2008b). In this approach, a
building Level of Protection is selected based on desired performance goals, as
shown in Table 3.3. Table 3.4 is then used to determine the associated damage
levels for each type of component in the building. Therefore, this approach determines component damage levels that are consistent with an overall building
performance goal. It can also be used to assess the Level of Protection provided
by an existing building, based on calculated component damage levels.
Table 3.4 refers to three types of components: primary, secondary, and
nonstructural. Primary components are members whose loss would affect a
90
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.3
Level of
Protection
Building Levels of Protection
Building Performance Goals
Overall Building Damage
I (Very low)
Collapse prevention: Surviving occupants
will likely be able to evacuate, but the
building is not reusable; contents may not
remain intact.
Damage is expected, up to the
onset of total collapse, but
progressive collapse is
unlikely.
II (Low)
Life safety: Surviving occupants will likely
be able to evacuate and then return only
temporarily; contents will likely remain
intact for retrieval.
Damage is expected, such that
the building is not likely to be
economically repairable, but
progressive collapse is
unlikely.
III (Medium)
Property preservation: Surviving occupants
may have to evacuate temporarily, but
will likely be able to return after cleanup
and repairs to resume operations;
contents will likely remain at least
partially functional, but may be impaired
for a time.
Damage is expected, but
building is expected to be
economically repairable, and
progressive collapse is
unlikely.
IV (High)
Continuous occupancy: All occupants will
likely be able to stay and maintain
operations without interruption; contents
will likely remain fully functional.
Only superficial damage is
expected.
number of other supported components and whose loss could potentially affect
the overall structural stability of a building area. Examples of primary structural
components include columns, girders, and any load-bearing structural components such as walls. Secondary components generally are supported by a primary
framing component and can fail without creating widespread structural damage.
Table 3.4
Expected Component Damage for Each Level of Protection
Component Damage Levels
Level of Protection
Primary Structural
Components
Secondary Structural
Components
Nonstructural
Components
I (Very low)
Heavy
Hazardous
Hazardous
II (Low)
Moderate
Heavy
Heavy
III (Medium)
Superficial
Moderate
Moderate
IV (High)
Superficial
Superficial
Superficial
Hazardous damage—The element is likely to fail and produce debris.
Heavy damage—The element is unlikely to fail, but will have significant permanent deflections such that it is not
repairable.
Moderate damage—The element is unlikely to fail, but will probably have some permanent deflection such that
it is repairable, although replacement may be preferable for economic or aesthetic reasons.
Superficial damage—The element is unlikely to exhibit any visible permanent damage.
RESPONSE PARAMETERS
91
Examples of secondary structural components include non-load-bearing infill
masonry walls, metal panels, and secondary steel framing members such as
studs, girts, purlins, and joists. Nonstructural components include interior
non-load-bearing walls, and architectural items attached to building structural
components.
Nonstructural components are not usually designed specifically against design
blast loads. In general, their performance has received very little attention, since
testing and research for blast-resistant design has focused on structural components. However, when appropriate, depending on the level of protection, these
nonstructural components can be attached to structural members with strengthened connections in manner similar to that used in earthquake-resistant design.
3.3 RESPONSE PARAMETERS
Component blast damage levels, such as those in Table 3.4, have been correlated
to the maximum dynamic deflection of the component with two nondimensional
parameters: the ductility ratio (µ) and support rotation (θ). The reasons for basing the correlations on these simplified parameters are discussed in this section,
as well as the limitations of this approach. The support rotation is defined in
Figure 3.1:
2xm
θ = tan−1
L min
where: Lmin = is the shortest distance from a point of maximum deflection to a
support
It is based on the shortest distance from a support to the point of maximum component deflection (Lmin ), where the point of maximum deflection is determined
from yield line theory. Lmin is equal to one-half the span length of one-way spanning components, except for cantilevers where Lmin is the whole span length.
The yield line pattern for two-way spanning components can be determined from
charts in UFC 3-340-02 (Unified Facilities Criteria Program 2008), which is an
update of TM 5-1300 (DOANAF 1990).
The ductility ratio is defined as the ratio of the maximum component deflection to the component yield deflection, as shown in Equation 3.1. The yield
L
Lmin
θ = Support Rotation
XM
Figure 3.1 Support Rotation Angle
92
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
rU
Ke
Kep
1
re
KE
RESISTANCE
Ke
RESISTANCE
rU
1
1
1
Xe
Xe
Xm
XE
Xp
Xm
DEFLECTION
DEFLECTION
Determinate Boundary Conditions
Determinate Boundary Conditions
Figure 3.2 Resistance-Deflection Curves for Flexural Response
deflection for determinate components is the deflection when the stress in the
maximum moment region equals the yield strength. The yield deflection for indeterminate components is usually equal to the equivalent yield deflection, XE ,
in Figure 3.2. This is an average ductility ratio, which is less than the ductility ratio at the initial yield location and greater than the ductility ratio at the
final yield location. It is calculated as shown in Equation 3.2, such that the area
under the resistance-deflection graphs for the simplified elastic-perfectly plastic
resistance-deflection curve yielding at XE (i.e., dashed curve in Figure 3.2) and
the actual, multi-stiffness resistance-deflection curve (i.e., solid curve in Figure
3.2) are equal at deflections greater than the deflection at which the component
becomes a mechanism (i.e., Xp in Figure 3.2).
µ=
Xm
Xy
where: µ = ductility ratio
Xm = maximum component deflection
Xy = deflection causing yield of component
= Xe for determinate components (see Figure 3.2)
= XE for indeterminate components (see Figure 3.2)
re
X E = Xe + X p 1 −
ru
ru
KE =
XE
(3.1)
(3.2)
Given values of ductility ratio and support rotation have traditionally been
used as response criteria in blast-resistant design, where these criteria are the
quantitative limits that define the upper bounds of each response level or damage
level for blast-loaded components. These two parameters are used as the basis for
response criteria because they are related to the amount of component damage,
as discussed below, and because of their design-level simplicity. They are both
RESPONSE PARAMETERS
93
based on the maximum dynamic component deflection, which is relatively easy
to measure in component blast tests and to calculate with simplified dynamic
analysis methods used for blast design.
Damage occurs in blast-loaded components as plastic strains approach the
material failure strain at some location in the component. The ductility ratio is
an approximate measure of component plastic strain based on the assumption
that the curvature in the maximum moment regions increases proportionally
with deflection after yielding, and plane cross sections remain plane. In this
case, the ductility ratio is a measure of the ratio of the total strain to the yield
strain in the maximum fibers. This approach is applicable to cases where the
strains causing the component damage occur at the same location where initial
yielding occurs, such as the extreme fiber of the maximum moment region of
a steel beam. In reinforced concrete, on the other hand, damage is primarily
controlled by compression strains as they approach a given limit value (i.e., the
concrete crushing strain), whereas initial yielding occurs in the tension steel
at a deflection that usually causes relatively low compression strains in the
maximum moment region. Therefore, the ductility ratio does not correlate well
with flexural damage to reinforced concrete members. Also, very ductile steel
members typically develop tension membrane response at high ductility levels,
which can causes high strains and failure around the connections, rather than at
midspan, where initial yielding occurs.
The support rotation correlates better to component blast damage than the
ductility ratio in these cases. In the case of reinforced concrete and masonry components, the support rotation is related to the cross section rotation at maximum
moment regions, and therefore to the compression face strain, which primarily
causes damage and failure to underreinforced components. This is the case when
the component responds as a ductile mechanism after yielding of the reinforcing
steel. Very little damage occurs prior to yielding of the reinforcing steel.
In the case of tension membrane response, the support rotation increases proportionally with the maximum deflection-to-span ratio and therefore increases
with the amount of tension membrane response. Support rotation limits help ensure that the amount of tension membrane force in a ductile component that
can undergo large deflections is limited to values that do not allow connection
failures. Large deflections of primary components can also cause failures at connections to supported components. Many blast tests and field investigations at
accidental explosion sites have shown that steel components connected to their
supports with screws, bolts, and welds typically fail at the connection, rather
than at the points of maximum strain in the component (i.e., yield locations).
Therefore, ductile steel components typically have separate response criteria in
terms of ductility ratio, to limit plastic strains at the maximum moment region,
and in terms of support rotation, to limit strains from tension membrane response
near connections or other critical areas (i.e., utility cutouts in well-attached metal
studs).
The ductility ratio is usually most useful for determining damage levels with
low ductility ratios in the range of 1.0 to approximately 3.0. This applies to low
94
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
damage levels for ductile components and low to high damage levels of relatively
brittle components, such as wood components. Damage from brittle, nonflexural
response modes, such as shear, buckling, or connection failure, can also be correlated to the ductility ratio if the yield deflection, Xy , in these cases equals the
maximum component deflection causing a critical shear, buckling, or connection
strain in the component and these strains increase proportionally with increased
component deflection.
Limit support rotations and ductility ratios that establish component damage
levels generally cannot be derived from theoretical considerations alone. These
parameters are simplified for practical reasons and failure modes are complex
and influenced significantly from actions, such as combined bending, shear, and
axial load, which are difficult if not impossible to quantify and characterize universally. Also, applicable damage levels typically have qualitative, rather than
quantitative, descriptions. Therefore, limit values for these parameters are based
primarily on empirical correlations between observed damage levels to components from blast tests and corresponding values of support rotations and ductility
ratios calculated from the component’s measured maximum dynamic deflection.
This is discussed more in the next section.
A limitation in this approach is that the support rotation and ductility ratio
terms do not capture all the parameters that are actually related to the observed
component blast damage. For example, an analysis of data from fifteen oneway-spanning reinforced masonry walls tested in a shock tube showed that the
relationship between the measured support rotations and the observed damage
levels was strongly dependent on the wall reinforcement factor, which compares
the reinforcing steel ratio to the balanced steel ratio, as well as to the support
rotation (Oswald and Nebuda 2006). The test walls with a low reinforcement
factor sustained significantly less damage at the same support rotation than walls
with a higher factor. Also, a statistical study of several hundred static reinforced
concrete beam and slab tests showed that the support rotation at failure was not
constant, as generally assumed for blast design, but was a function of several
parameters, including the reinforcing steel ratio, the shear span to depth ratio,
and the ratio of the steel reinforcement yield strength to concrete compression
strength (Panagiotakos and Fardis 2001). Therefore, if two concrete panels with
different spans, thicknesses, and reinforcement ratios are both subject to blast
loads causing the same given support rotation (e.g., 2 degrees), it is possible that
they will not have the same observed damage level. A similar situation is possible
for other component types. Fortunately, this is alleviated to some extent by the
qualitative nature of damage levels, which encompass relatively broad ranges of
damage.
In spite of the limitations of the support rotation and ductility ratio as
definitive measures of component damage, it is the judgment and experience
of engineers who have developed response criteria relating these parameters to
component damage that they are adequate and generally conservative for blast
design and analysis purposes. More accurate terms that include more properties associated with a component’s blast damage level, and have acceptable
simplicity for design-based methodologies, may be developed in the future.
RESPONSE PARAMETERS AND DAMAGE CORRELATIONS
95
3.4 EMPIRICAL CORRELATIONS BETWEEN RESPONSE
PARAMETERS AND COMPONENT DAMAGE
Support rotations and/or ductility ratios that define the damage levels of blastloaded components have been established using correlations between these two
parameters calculated from available test data and observed damage levels in
the test components. The amount of available test data varies with component
type, since blast testing is typically conducted by various government agencies
and organizations to address specific blast design considerations, and the data
include varying amounts of measured material properties, blast load information,
and response values. Also, there is an almost complete lack of blast test data
for some component types and damage levels. Deflection-controlled static tests
that provide information on component performance at post-yield deflections
are considered for component types with limited available blast test data. These
static data are generally conservative because each strain and deformation level
exists for a sustained period of time, and can therefore result in more damage
than the same strain level that will only exist from a blast load for a few tenths
of a second, or less.
The most comprehensive published study that has correlated blast test data
with component damage levels is part of the Component Explosive Damage Assessment Workbook (CEDAW) methodology (Protective Design Center 2008a).
CEDAW is a fast-running methodology to assess component blast damage based
on pressure-impulse (P-I) diagrams that was developed to be as consistent as
possible with available blast test data. In this study, data were collected from
over 300 blast tests on 10 different component types and used to calculate the
support rotation and ductility ratio for each test, using single-degree-of-freedom
(SDOF) dynamic analyses. Then, this information was used to determine the rotations and ductility ratios that bounded each observed damage level for each
component type, using scaled P-I diagrams.
This effort is illustrated in Figure 3.3 and Figure 3.4, which are based on
damage levels defined in Table 3.5. Pbar and Ibar in the graphs are the peak
positive phase pressure and impulse, respectively, scaled (i.e., divided) by
relevant component dynamic response parameters to create dimensionless terms
that are directly related to component response. Therefore, components with
very different blast loads and structural properties, but the same Pbar and Ibar
values, should have the same ductility ratio or support rotation values (i.e.,
a strong component with a high blast load and a weaker component with a
lower blast load that both have response with the same ductility ratio or support
rotation would have the same Pbar and Ibar values). This should be true in both
blast tests and dynamic response calculations. The Ibar terms for component
response in terms of ductility ratio are different from those for component
response in terms of support rotation.
Figure 3.3 shows a scaled P-I diagram with curves of scaled blast loads
calculated with SDOF analyses causing ductility ratios (µ) of 1, 3, 6, and 12 for
a corrugated steel panel. These curves are plotted with data points of scaled blast
loads causing damage levels ranging from Superficial to Hazardous Failure to
96
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
100
Superficial
Moderate
Heavy
Haz. Failure
µ=1
µ=3
10
Sup. Data
Mod. Data
Heavy Data
H.Fail Data
µ=6
Pbar
µ = 12
1
0.1
0.1
1
10
100
Ibar 1
Figure 3.3 Scaled P-I Curves in Terms of Ductility Ratio vs. Scaled Data for Flexural
Response of Corrugated Steel Panels
corrugated steel panels in blast tests. Figure 3.4 is a similar scaled P-I diagram
with curves of scaled blast loads calculated with SDOF analyses causing support
rotations (θ) of 3, 6, and 10 degrees plotted against the same data that are
scaled by Pbar and Ibar terms based on component support rotation. In both
of these figures, the curves with the lowest support rotation or ductility ratio
correspond to the lowest damage level (i.e., Superficial), and curves with higher
support rotations and ductility ratios correspond to higher damage levels up to
Hazardous Failure.
Figure 3.3 and Figure 3.4 were used to determine that the ductility ratios and support rotations shown on the figures bounded each of the damage
Pbar
100
Moderate
Heavy
Haz. Failure
Mod. Data
Heavy Data
H.Fail Data
Blowout Data
θ=3
θ=6
10
θ = 10
1
0.01
0.1
0.1
1
Ibar 2
Figure 3.4 Scaled P-I Curves in Terms of Support Rotation (Degrees) vs. Scaled Data
for Flexural Response of Corrugated Steel Panels
RESPONSE PARAMETERS AND DAMAGE CORRELATIONS
97
Table 3.5 Component Damage-Level Descriptions
Damage Level
Component Damage
Blowout
Component is overwhelmed by the blast load, causing debris with
significant velocities.
Hazardous
Failure
Component has failed, and debris velocities range from insignificant to very
significant.
Heavy
Component has not failed, but it has significant permanent deflections,
causing it to be unrepairable.
Moderate
Component has some permanent deflection. It is generally repairable, if
necessary, although replacement may be more economical and aesthetic.
Superficial
No visible permanent damage.
levels, and therefore could be used as response criteria for corrugated steel panels (Protective Design Center 2008a). Similar curves were generated for all other
component types where there were available test data. The CEDAW Workbook
(Protective Design Center 2008a) has figures similar to Figure 3.3 and Figure
3.4 for each component type and a full description of the test data. Table 3.6
shows a summary of the response criteria determined for each component type
that was developed in this manner. No criteria are shown for Superficial Damage,
since this damage level was correlated to a ductility ratio of 1 by definition for
the CEDAW study. No test data were available to develop response criteria for
component types not shown in Table 3.6, including hot-rolled steel beams and
primary framing components.
PDC TR-08-07 (Protective Design Center 2008a) and Oswald and Nebuda
(2006) have a detailed discussion of the Pbar and Ibar terms used to develop
the empirical correlations between component performance levels and support
rotations and ductility ratios. These terms were developed for component
response in terms of support rotation and ductility ratio and for different types
of response modes (i.e., flexure, tension membrane, etc.) based on conservation
of energy equations between the work energy and kinetic energy from the blast
load and the strain energy of the component in the applicable response mode.
The parameters included in the Pbar and Ibar terms vary based on the type of response mode for the component and the type of response parameter representing
the component response (i.e., ductility ratio or support rotation). Also, the Pbar
and Ibar terms were developed to account for both positive and negative phase
blast loads since both types of blast loads generally affect component response
and damage levels. Extensive checks were performed on the validity of the Pbar
and Ibar scaling terms, as summarized in PDC TR-08-07 (Protective Design
Center 2008a). The scaling terms are not exact since they were derived based
on some simplifying approximations, but numerous comparisons showed they
caused the SDOF-based scaled P-I diagrams to be within 20% of more exact
solutions in almost all cases.
98
8
4
2
1.5
One-way reinforced masonry
One-way or two-way
unreinforced masonry1
4
5
2
Reinforced concrete beam
Response criteria are only valid for SDOF analyses with brittle flexural response and axial load arching.
1
Steel column (connection
failure)
1
6
Reinforced concrete column
(shear failure)
Wood stud wall
5
2
6
Heavy
One-way or two-way
reinforced concrete slab
3
Moderate
6
3
12
Haz Failure
3
2
6
Heavy
From data
w/o
SDOF
15
10
10
10
10
Haz Failure
Support Rotation
Open-web steel joist
Cold-formed girt and purlins
3
Moderate
Ductility Ratio
Response Criteria for Upper Bound of Each Damage Level for Each Component Type
One-way corrugated metal
panel
Component
Table 3.6
3
4
Moderate
6
12
Heavy
10
20
Haz Failure
Support Rotation w/ Tension
Membrane
RESPONSE CRITERIA DEVELOPMENT
99
3.5 RESPONSE CRITERIA DEVELOPMENT
The information in Table 3.6 represents the most extensive and formalized use
of blast test data to develop response criteria. This information was an important
part of the development of several of the current blast response criteria, However,
it was not the only consideration. Each organization that has developed response
criteria defining given performance levels considers a range of factors, such as
the exact definitions of its performance levels relative to the available data, how
to address component types where the data are inadequate, whether to discount
some data or data sets as not applicable, and the level of conservatism to be
included in the response criteria. There is very limited published information on
these considerations or on the compilations of the data that were considered.
Table 3.7 shows a summary of response criteria that have been developed for
each area of blast design. Ideally, there would be one general set of response limits on component response applicable for all the performance levels and design
strategies in Table 3.1. There is strong movement in this direction, but currently
there are separate sets of response limits associated with performance levels established by the different organizations associated with each area of blast design.
These differences are due in part to the fact that these organizations have different missions, design objectives, and types of explosive threats.
The following subsections provide some detail on the criteria in Table 3.7. In
all cases the response criteria are stated in terms of either ductility ratio, support
rotation, or both terms, as discussed in Section 3.2. In cases where response criteria for both terms are provided, the overall component damage level or performance level is controlled by the response criterion that causes more severe damage or lower performance. In addition to the response criteria discussed below,
the engineer should also always consider any design-case-specific requirements
on component deflection and limitations of response criteria, as discussed in Section 3.6. It should also be noted that all the response limits stated in this document
may be updated in the future. Historically, response criteria have been updated
as more information on component response to blast loads has become available
and as new component and material types are used for blast-resistant design.
3.5.1 Explosive Safety Criteria
Blast-resistant design criteria are included in UFC 3-340-02 (Unified Facilities
Criteria Program 2008), which has been developed by the explosive safety design
community under direction of the U.S. Department of Defense Explosive Safety
Board (DDESB). Explosive safety refers to safe design and operation of manufacturing and storage areas for high explosives that can accidently explode. This
primarily includes military and military contractor facilities with explosives, but
it also includes industries such as fireworks and commercial explosives manufacturers.
Most blast-resistant design prior to 1990 was for explosive safety and critical military structures, and the majority of the initial blast testing of structural
100
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.7
Summary of Response Criteria Publications
Organization
Publication
Type of Criteria
Comments
Department of
Defense
UFC 3-340-02
(Unified
Facilities
Criteria
Program 2008)
Design criteria for
explosive safety.
Published in draft form in 1984
and finalized in 1990 as the
tri-service manual TM
5-1300, NAVFAC P397, and
AFM 88 22. This document
generally has the most
conservative criteria. It has
been updated recently with
some revisions as UFC
3-340-02.
Task Committee
on Blast
Resistant
Design of the
Petrochemical
Committee of
the Energy
Division of
ASCE
Initial performance criteria for
Performance criteria
Design of Blast
blast design from 1997.
for three component
Resistant
These criteria will be updated
damage levels.
Buildings in
in a new edition of this
These criteria are
Petrochemical
publication in 2009.
commonly used for
Facilities
blast design for
(American
industry related to
Society of Civil
process safety.
Engineers
1997b)
U.S. Army Corps
of Engineers,
Protective
Design Center
(PDC)
Single-Degree-ofFreedom
Structural
Response
Limits for
Antiterrorism
Design, PDC
TR-06-08
(Protective
Design Center
2008b).
ASCE Blast
Protection of
Buildings
Standards
Committee
Based very strongly on criteria
Blast Protection of Performance criteria
in PDC TR-06-08 from U.S.
for four component
Buildings
Army Corps of Engineers.
damage levels.
(American
These criteria can be
Society of Civil
used for a wide
Engineers,
range of blast design
forthcoming)
except explosive
safety.
Performance criteria
for four component
damage levels
associated with
building levels of
protection. These
criteria are
commonly used for
design against
terrorist threats.
Performance criteria developed
in 2006 that are significantly
more comprehensive than
previous criteria. Also, they
are based more explicitly on
available blast test data.
components against conventional (i.e. non-nuclear) explosives was conducted in
support of explosive safety design. Since that time, an increase in terrorism and
industrial safety requirements by OSHA have considerably increased the amount
of blast-resistant design.
Response criteria for blast resistant design related to explosive safety for the
U.S. Department of Defense are defined in UFC 3-340-02 for the component
types shown in Table 3.8. Inhabited buildings subject to blast pressures from
RESPONSE CRITERIA DEVELOPMENT
Table 3.8
101
Response Criteria for Explosive Safety Design from UFC 3-340-02
Protection Categorya
1
2
µMAX
MAX
µMAX
MAX
Reinforced concreteb
Conventional slabs
Beams with stirrups (closed ties)
Exterior columns
—
—
3
1◦c
2◦
—
—
—
—
8◦
8◦
—
Precast concrete
Prestressed
Non-prestressed
Compression members
1
3
1.3
2
2
—
—
—
—
—
—
Reinforced masonryb
One-wayd
Two-wayd
—
—
1◦
2◦
—
—
10
—
10
2◦
2◦e
2◦
20
—
20
12◦
—
12◦
Open-web steel joists
Controlled by maximum end reaction
Otherwise
1
4
1◦
2◦
—
—
—
—
Cold-formed steel floor and wall panels
Without tension membrane action
With tension membrane action
1.75
6
1.25◦
4◦
—
—
—
—
Element Type
Structural steel
Beams, purlins, spandrels, girts
Frame structures
Plates
—
Note: Where a dash (—) is shown, the corresponding parameter is not applicable as a response limit.
a
See Table 3.9 for definitions of Protection Categories.
b
Reinforced with conventional steel rebar.
c
May be increased to 2◦ with either shear reinforcement or tensile membrane action.
d
Non-reusable walls.
e
Relative sidesway deflection between stories is limited to H/25.
accidental explosions must resist the blast loads with response criteria for Protection Category 1 in Table 3.9. Buildings protecting critical equipment must
resist the blast loads with response criteria for Protection Category 2. The component damage associated with each Protection Category is not well defined in
UFC 3-340-02, but Protection Category 1 is considered to provide a factor of
safety against failure of at least two.
The performance criteria in Table 3.8 were first published in 1984 in a draft
version of TM 5-1300 and then finalized in 1990. They are unchanged in the
recent conversion of TM 5-1300 into UFC 3-340-02. These criteria are more
conservative than more recent criteria discussed in the next sections, which are
based in part on many recent blast tests on structural components that were not
available at the time the response criteria for TM 5-1300 were developed. It
102
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.9 Protection Category Descriptions for Structural Components from
UFC 3-340-02
Protection Category
Description for Structural Components
Category 1
Attenuate blast pressures and structural motion to a level consistent
with personnel tolerances
Category 2
Protect equipment, supplies, and stored explosives from fragment
impact, blast pressures, and structural response
is also the intent of TM 5-1300/ UFC 3-340-02 to provide design criteria with
a significant level of conservatism, since it is used to help protect workers at
explosive production and storage facilities, who spend a significant amount of
time exposed to the risk of accidental explosions.
3.5.2 Response Criteria for Antiterrorism
The Protective Design Center (PDC) of the U.S. Army Corps of Engineers published Single-Degree-of-Freedom Structural Response Limits for Antiterrorism
Design in 2006 (i.e. DoD Antiterrorism response criteria) and updated it in
2008 (Protective Design Center 2008b). These response criteria have also been
adopted into the forthcoming ASCE document “Blast Protection of Buildings”
(American Society of Civil Engineers, forthcoming). The PDC response criteria
define limit values for ductility ratio and support rotation corresponding to four
different damage levels that apply to twelve different common structural component types with a total of 38 different subcategories based on component characteristics, such as response mode and types of connections. They also consider
the cases where components are primary components (i.e., major framing components) or secondary components (i.e., cladding components). These response
criteria are used to design DoD facilities against explosive terrorist threats and
can also be used for non-DoD government buildings, such as those that provide
blast protection required by the Interagency Security Committee (ISC) Security
Criteria (Interagency Security Committee 2004).
The DoD Antiterrorism response criteria are shown in Table 3.10. Some of
the criteria are based on the response limits developed from blast test data acquired to develop CEDAW (shown in Table 3.6) for corresponding component
types. The response criteria for other component types are based primarily on
static test data, as summarized in Table 3.11. The response criteria based on static
test data are generally more conservative than those based on blast test data, as
discussed previously. For example, there are limited blast test data that indicate
that conventional metal stud wall systems can resist significantly higher loads
than those calculated with SDOF analysis using the response limits in Table 3.10
(Grumbach et al. 2007).
103
—
1
1
1
1
0.7
4
1
1
0.7
1
0.5
Masonry
Unreinforcedc,e
Reinforceda
Structural steel (hot-rolled)
Beam with compact sectionf
Beam with noncompact sectionc,f
Plate bent about weak axis
Open-web steel joist
Downward loadingg
Upward loadingh
Shear responsei
Cold-formed steel
Girt or purlin
Stud with sliding connection at top
—
—
—
—
—
—
—
1◦
—
—
—
—
—
—
1
0.7
0.8
0.8
—
—
MAX
1
1
µMAX
Prestressed concrete
Slab or beam with ωp > 0.30
Slab or beam with 0.15 ≤ ωp ≤ 0.30
Slab or beam with ωp < 0.15 and without shear
reinforcementb,c
Slab or beam with ωp < 0.15 and shear reinforcementb
d
Reinforced concretea
Single-reinforced slab or beam
Double-reinforced slab or beam without shear
reinforcementb,c
Double-reinforced slab or beam with shear reinforcementb
Element Type
Superficial
Table 3.10 DoD Response Criteria for Antiterrorism Design
—
0.8
—
1.5
0.8
3
0.85
8
—
—
—
0.8
0.25/ωp
0.25/ωp
—
—
—
µMAX
Heavy
3◦
—
3◦
—
—
2◦
3◦
—
—
0.9
2
0.9
12
1
20
—
—
—
1◦
1.5◦
2◦
0.9
0.29/ωp
0.29/ωp
—
4◦
—
1◦
1◦
—
—
µMAX
2◦
2◦
MAX
Moderate
10◦
—
6◦
—
—
10◦
—
6◦
4◦
8◦
2◦
—
1.5◦
1.5◦
6◦
5◦
5◦
MAX
Expected Component Damage
1
3
1
—
—
25
1.2
40
—
—
—
1
0.33/ωp
0.33/ωp
—
—
—
µMAX
(Continued)
20◦
—
10◦
—
—
20◦
—
12◦
8◦
15◦
3◦
—
2◦
2◦
10◦
10◦
10◦
MAX
Hazardous
104
j
3
3
Blast doors
Built-up (composite plate & stiffeners)
Plate (solid)
1
1
—
—
—
—
—
—
10
20
2
1
1
3
—
1.8
µMAX
6
6
—
—
0.5◦
3◦
1◦
1.3◦
MAX
Moderate
20
40
3
2
2
6
—
3
µMAX
12
12
—
—
2◦
6◦
4◦
2◦
MAX
Heavy
Expected Component Damage
—
—
4
3
5
10
—
6
µMAX
—
—
—
—
5◦
12◦
8◦
4◦
MAX
Hazardous
Note: Where a dash (—) is shown, the corresponding parameter is not applicable as a response limit.
a
Reinforced with conventional steel rebar.
b
Stirrups or ties that satisfy sections 11.5.5 and 11.5.6 of ACI 318 and enclose both layers of flexural reinforcement throughout the span length.
c
The ASCE” Blast Protection of Buildings” document states that these response limits are applicable for flexural evaluation of existing elements that satisfy the design requirements
of Chapters 6 through 8 but do not satisfy the detailing requirements in Chapter 9 of the document, and shall not be used for design of new elements. This note does not appear in
the DoD Response Criteria for Antiterrorism Design.
d
Span to thickness ratio greater than 5. Also, reinforcement index ωp = (Aps /bd)(fps /f c ). Also see Equation 3.3.
e
Values assume wall resistance controlled by brittle flexural response or axial load arching with no plastic deformation; for load-bearing walls, use superficial or moderate damage
limits to preclude collapse.
f
Limiting width-to-thickness ratios for compact and noncompact sections are defined in ANSI/AISC 360.
g
Values assume tension yielding of bottom chord with adequate bracing of top chord to prevent lateral buckling.
h
Values assume adequate anchorage to prevent pull-out failure and adequate bracing of bottom chord to prevent lateral buckling.
i
Applicable when element capacity is controlled by web members, web connections, or support connections; ductility ratio for shear is equal to peak shear force divided by shear
capacity.
j
Also applicable when studs are continuous across a support.
k
Requires structural plate-and-angle bolted connections at top, bottom, and any intermediate supports.
l
Panel has adequate connections to yield cross section fully.
m
Typically applicable for simple-fixed span conditions.
n
Limited to connector capacity; includes all standing seam metal roof systems.
o
Values shown are based on very limited testing data; use specific test data if available.
1
0.5
0.5
1
1
1
MAX
Superficial
µMAX
Wood◦
Stud connected at top and bottom
Stud with tension membranek
Corrugated panel (1-way)with full tension membranel
Corrugated panel (1-way) with some tension membranem
Corrugated panel (1-way) with limited tension membranen
Element Type
Table 3.10 (Continued)
RESPONSE CRITERIA DEVELOPMENT
Table 3.11
105
Bases for DoD Antiterrorism Response Criteria
Component Type
Primary Basis for Response Criteria
Reinforced concrete slabs
Response limits from CEDAW study based on several blast
test series on simply supported wall panels where damage
occurred primarily in maximum moment region.
Reinforced concrete beams
Blast test data for slabs assumed to apply to reinforced
concrete beams.
Hot-rolled steel beams
Static test data.
Cold-formed steel panels and
girts/purlins
Response limits from CEDAW study based on several blast
test series on wall systems with cold-formed beams and
corrugated steel panels where lower damage levels
occurred primarily in maximum moment region,
characterized by local buckling of compression flange, and
higher damage levels including failure occurred at the
connections due primarily to in-plane reaction forces from
tension membrane response that occurs at large deflections.
Cold-formed metal studs
Static test data.
Open-web steel joists
Downward response limits from CEDAW study based on one
test series.
Wood
Response limits from CEDAW study on two test programs on
wood stud walls. One static and one dynamic test series
both indicated that typical plywood panels did not create a
fully composite section that added significantly to ultimate
resistance of stud wall system.
Prestresssed concrete
components
Static test data.
Reinforced masonry
Several test programs with shock tube and high-explosive
tests.
Unreinforced masonry
Several test programs with shock tube and high-explosive
tests documented in PDC TR-08-07 (2008a) and Wesevich
et al. (2002).
3.5.3 Response Criteria for Blast-Resistant Design of
Petrochemical Facilities
The Task Committee on Blast Resistant Design of the Petrochemical Committee
of the Energy Division of ASCE published the manual Design of Blast Resistant
Buildings in Petrochemical Facilities in 1997 (American Society of Civil Engineers 1997b), and an update will be finalized in 2009. This manual has three levels of component response to blast loads, as shown in Table 3.12, and provides
response criteria in terms of maximum ductility ratio and/or support rotation
criteria for each response level. Separate criteria are defined for different component types (i.e., steel beams, concrete slabs, etc.), and the criteria address primary and secondary components, frame components with and without significant
axial load, and shear controlled response for reinforced concrete components.
106
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.12 Component Damage Level Descriptions
Damage Level
Description
High
Component has not failed, but it has significant permanent deflections,
causing it to be unrepairable.
Medium
Component has some permanent deflection. It is generally repairable, if
necessary, although replacement may be more economical and
aesthetic.
Low
Component has none to slight visible permanent damage.
In general, the response limit values are quite low for components in which shear
or compression is significant compared to flexural response. Where adequate
shear capacity is provided, much larger response limit values are associated with
each response or damage level.
The updated version contains new response criteria that are based on a combination of criteria developed for the original publication and the response criteria
in PDC TR-06-08. The intent for the new response criteria was to expand the
existing criteria to consider a wider range of component types and subcategories
relevant for typical industrial buildings and to replace or modify existing response criteria with values from PDC TR-06-08 where there was a strong case to
be made based on the CEDAW data (Oswald 2008). Also, differences in damage
level definitions used by the PDC and by the ASCE committee (see Table 3.12)
were considered.
Table 3.13 shows the updated response limits for steel components, and Table
3.14 shows the updated response criteria for reinforced concrete and masonry
components. The criteria values for Medium and High response of cold-formed
steel girts and purlins, reinforced concrete and masonry, and prestressed concrete
components in Table 3.13 and Table 3.14 are based on similar values for Moderate and Heavy component damage levels, respectively, in PDC TR-06-08. The
criteria for Superficial damage in PDC TR-06-08 were not used in the ASCE
response criteria because this damage level was considered more conservative
than the Low damage level in Table 3.12. Therefore, values for Low response
of these component types are approximately one-half the values for Medium response. The reinforcement index that is used as part of the response criteria for
prestressed concrete components in Table 3.14 is defined in Equation 3.3.
The response criteria for all other component types not listed in the previous
paragraph are the same, or very similar, to values developed for the original
publication of the manual. In some cases this is because the values in PDC TR06-08 are similar to those in the original publication of the manual or there were
not enough supportive blast test data to justify changes.
wp =
A ps f ps
bd p f c
(3.3)
107
RESPONSE CRITERIA DEVELOPMENT
Table 3.13 Response Criteria for Steel Components
Low Response
µMAX
Component1
θ MAX
Medium
Response
µMAX
θ MAX
µMAX
θ MAX
20
12
Hot-rolled steel compact secondary
members (beams, girts, purlins)2
3
2
10
Steel primary frame members (with
significant compression)2,3,4
1.5
1
2
1.5
3
2
Steel primary frame members
(without significant
compression)2,3,4
1.5
1
3
2
6
4
Steel plates5
5
3
10
6
20
12
Open-web steel joists
1
1
2
3
4
6
Cold-formed light gage steel panels
(with secured ends)7
1.75
1.25
3
2
6
4
Cold-formed light gage6 steel panels
(with unsecured ends)8
1.0
—
1.8
1.3
3
2
Cold-formed light gage6 steel beams,
girts, purlins, and noncompact
secondary hot-rolled members
2
3
3
12
10
6
1.5
6
High Response
1
Response limits are for components responding primarily in flexure unless otherwise noted. Flexure controls
when shear resistance is at least 120% of flexural capacity.
2
Primary members are components whose loss would affect a number of other components supported by that
member and whose loss could potentially affect the overall structural stability of the building in the area of loss.
Secondary members are those supported by primary framing components.
3
Significant compression occurs when the axial compressive load, P, is more than 20% of the dynamic axial
capacity of the member. P should be based on the ultimate resistance of the supported members exposed to the
blast pressure. See PDC TR-06-08 for detailed examples of calculation for P.
4
Sidesway limits for moment-resisting structural steel frames: low = (height)/50, medium = (height)/35, high =
(height)/25.
5
Steel plate criteria can also be applied to corrugated (crimped) plates if local buckling and other response modes
are accounted for in the analysis.
6
Light gage refers to material which is less than 0.125 inches (3 mm) thick.
7
Panels must be attached on both ends with screws or spot welds.
8
Panels are not attached on both ends (for example standing seam roof panels).
where: Aps = area of prestressed reinforcement in tension zone
b = the member width
dp = the depth from compression face to center of prestressing steel
fps = calculated stress in prestressing steel at design load
f c = the concrete compressive strength
3.5.4 Blast Resistant Doors
Table 3.15 shows recommended response criteria for design of structural components in blast doors based on a desired door performance category from Design
108
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.14 Response Criteria for Reinforced Concrete (R/C) and Masonry (R/M)
Components
Medium
Response
Low Response
Component1
µMAX
θ MAX
µMAX
High Response
θ MAX
µMAX
θ MAX
R/C beams, slabs, and wall panels (no
shear reinforcement)
1
2
5
R/C beams, slabs, and wall panels
(compression face steel reinforcement
and shear reinforcement in maximum
moment areas)
2
4
6
Reinforced masonry
1
2
5
2
1
R/C walls, slabs, and columns (in flexure
and axial compression load)3
22
2
R/C and R/M shear walls & diaphragms
3
3
3
R/C and R/M components (shear control,
without shear reinforcement)
1.3
1.3
1.3
R/C and R/M components (shear control,
with shear reinforcement)
1.6
1.6
1.6
Prestressed concrete (wp ≤ 0.15)4
1
Prestressed concrete (0.15 < wp < 0.3)4
1
1
0.25/wp
1
2
0.29/wp
1.5
1
Response limits are for components reinforced with conventional rebar or prestressing strands responding primarily in flexure unless otherwise noted.
2
A support rotation of 4 degrees is allowed for R/C components that have compression face steel reinforcement
and shear reinforcement in maximum moment areas.
3
Applicable when the axial compressive load, P, is more than 20% of the dynamic axial capacity of the member.
P should be based on the ultimate resistance of the supported members exposed to the blast pressure. Refer to
PDC TR-06-08 for detailed examples of calculation for P.
4
The reinforcement index, wp , is defined in Equation 3.3.
of Blast Resistant Buildings in Petrochemical Facilities (American Society of
Civil Engineers 1997b). Table 3.10 shows similar design information from PDC
TR-06-08. Blast doors can be designed using SDOF analyses of door components responding in flexure such that the responses of all components comply
with the applicable response criteria in these tables. Both inbound and rebound
response of blast doors must typically be considered during blast door design.
The structural performance of metal doors and frames and their restraining hardware (such as latches and hinges) at the load equal to the maximum calculated
resistance from the SDOF analyses can be verified by applying an equivalent
static pressure to blast doors in accordance with ASTM F2247 (ASTM 2003b).
Blast-resistant doors are typically designed by vendors according to specifications from the design engineer and architect. Blast door performance can also
be demonstrated by blast tests. There is an ASTM committee that has developed
a draft specification for blast testing of doors (i.e., ASTM WK1902).
109
RESPONSE CRITERIA DEVELOPMENT
Table 3.15 Recommended Response Criteria for Blast Door Components
Performance
Category
Response Criteria
Description
µMAX
θ MAX
I
The door is to be operable after the loading
event, and preestablished design criteria for
stress, deflection, and the limitation of
permanent deformation have not been
exceeded. This category should be specified
when the door may be required to withstand
repeated blasts or when entrapment of
personnel is of concern and the door is a
primary exit to the building.
1.0
1.2
II
The door is to be operable after the loading
event, but significant permanent deformation
to the door is permitted. The door must
remain operable, and this category should be
specified when entrapment of personnel is a
concern
2–3
2.0
III
Noncatastrophic failure is permitted. The door
assembly remains in the opening. No major
structural failure occurs in the door panel
structure, the restraining hardware system, the
frame, or the frame anchorage that would
prevent the door assembly from providing a
barrier to blast wave propagation. However,
the door will be rendered inoperable. This
category should only be specified when
entrapment of personnel is not a possibility.
5–10
8.0
IV
Outward rebound force and resulting hardware
failure is acceptable.
Not specified
Not specified
3.5.5 Blast-Resistant Windows
Table 3.16 shows performance criteria in the form of hazard levels to building
occupants for blast-loaded windows developed in the United Kingdom and used
by the U.S. DoD (SAFE/SSG 1997). Table 3.17 shows similar criteria developed by the Interagency Security Committee (ISC 2004). These criteria are illustrated in Figure 3.5 and Figure 3.6. Response of windows to blast load is currently calculated with government-sponsored software that is not public domain,
but is available from the U.S. government for design of U.S. government buildings on a need-to-know basis. These software programs, which include Wingard
(Window Glazing Analysis Response and Design) (Applied Research Associates 2008) and HazL (Window Fragment Hazard Level Analysis) (U.S. Army
Corps of Engineers 2004), model the overall window response using an equivalent SDOF system and glass flyout after window failure for a wide variety of
110
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.16
DoD Hazard Levels for Blast-Loaded Windows
Hazard Level
Description
No break
No visible damage to the glazing or frame.
Minimal hazard
Glazing fragments inside the test structure are within a maximum
distance of one meter from the window line. For laminated windows or
windows with a film that is attached to the window frame, the window
may be cracked and fully retained by the frame, or it may leave the
frame and land outside the structure or inside the structure within 1
meter of the window opening.
Low hazard
Glazing fragments are thrown into the room for a distance of 1 to 3
meters, but do not exceed a height of 0.5 meters above the floor at the
3-meter distance. Injuries would be limited to lower body cuts, and
fatalities would not be expected, although there would be some risk to
persons within 1 to 2 meters of windows.
High hazard
Glazing fragments are thrown at high velocity into the occupied space
and impact the vertical surface at 3 meters behind the window above a
0.5 meter height. Serious injuries, including cuts to the upper body and
face from the flying fragments, would be expected. Fatalities could
occur.
Table 3.17 ISC Performance Conditions and Hazard Levels for
Blast-Loaded Windows
Performance
Condition
Protection
Level
Hazard
Level
1
Safe
None
Glazing does not break. No visible damage to
glazing or frame.
2
Very high
None
Glazing cracks but is retained by the frame.
Dusting or very small fragments near sill or on
floor acceptable.
3a
High
Very low
Glazing cracks. Fragments enter space and land
on floor no further than 1 meter from the
window.
3b
High
Low
Glazing cracks. Fragments enter space and land
on floor no further than 3 meters from the
window.
4
Medium
Medium
Glazing cracks. Fragments enter space and land
on floor and impact a vertical witness panel at a
distance of no more than 3 meters from the
window at a height no greater than 0.6 meter
above the floor.
5
Low
High
Glazing cracks and window system fails
catastrophically. Fragments enter space
impacting a vertical witness panel at a distance
of no more than 3 meters from the window at a
height greater than 0.6 meter above the floor.
Description of Window Glazing Response
RESPONSE CRITERIA DEVELOPMENT
111
Blast
HHHigh
Hazard
Zone
Window
High
Hazard
Threshold
No Break
No Hazard
Minimal Hazard
0.5 m
1.0 m
VLH – Very
Low Hazard
Zone
2.0 m
LH – Low
Hazard
Zone
Low Hazard
Threshold
Figure 3.5 Illustration of DoD Window Hazard Levels for Blast-Loaded Windows
(from ASTM 1642)
common window configurations. If the aforementioned software is not available, blast-resistant windows can be designed using an equivalent static design
approach in accordance with ASTM F2248 (ASTM 2003a) and E1300 (ASTM
2007) for a medium or low protection level. This procedure is also discussed by
Norville and Conrath (Norville and Conrath 2006).
1, 2
5
4
3a
3b
≤ 2 ft
≤ 3.3 ft
≤ 10 ft
Test window should be in the design
position or centered on the wall.
Figure 3.6 Illustration of ISC Window Hazard Levels for Blast-Loaded Windows (from
ISC, 2004)
112
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Blast-resistant windows are typically designed by vendors according to specifications from the design engineer and architect. The specifications include the
design blast load history and information on the required performance criteria.
3.5.6 Response Criteria for Equivalent Static Loads
Previously discussed response criteria based on support rotations and ductility
ratios apply for the typical blast design case where the component is designed
using a procedure that explicitly considers the dynamic component response,
such as an SDOF-based methodology or dynamic finite element analysis. Some
stiff components that are not directly loaded by blast, such as connections and
primary framing members in pure axial loading, have a short response time
compared to the expected rise time of the dynamic reaction load transferred by
supported members, and are usually designed using an equivalent static load
approach. The load capacity of the stiff component or connection must exceed
the equivalent static reaction load in these cases. If the dynamic analysis shows
the supported member does not yield, the equivalent static reaction load may be
based on the maximum elastic resistance of this member rather than the ultimate
resistance. However, this approach should be used with caution, since it will be
nonconservative if the actual blast load exceeds the design blast load.
The load capacity of the stiff component or connection is usually calculated
using applicable strength reduction factors from LRFD (load and resistance factor design) (i.e., ∅ factors), although sometimes a ∅ factor of 1.0 is used at the
discretion of the design engineer. Also, the load capacity can include a dynamic
increase factor on the yield strength, but this is typically only 1.05 for highstrength components (i.e., A325 and A490 bolts) and should be applied with
caution for cases where the capacity is controlled by a non-ductile response mode
(i.e., axial buckling or anchor bolt pullout from concrete). The equivalent static
reaction load is not multiplied by a load factor.
The equivalent static load approach has conservatism based on the use of a ∅
factor less than 1.0 (if applicable), the use of a design load equal to the theoretical
ultimate load that can be delivered by the supported component, and the fact that
a static approach is used for blast design where the maximum strains are only
applied for a few tenths of a second followed by load reversal. However, it is not
conservative if the reaction load is applied fast enough by the supported component compared to the response time of the “stiff” supporting component (i.e.
primary framing member) such that dynamic deflections occur greater than the
component’s failure deflection. This is not a concern for connections since they
are very stiff and respond quickly compared to the application of the reaction
load. Also, any component, including primary framing components, can resist
the applied dynamic load regardless of the load rise time if the component has
a load capacity greater than 1.2 times the peak dynamic load, and it can achieve
at least a ductility ratio of 3.0 without failing, based on SDOF response charts
in UFC 3-340-02 (UFC, 2008), ASCE (American Society of Civil Engineers
1997b), and Biggs (1964).
RESPONSE CRITERIA DEVELOPMENT
113
Table 3.18 Comparison of Response Criteria for Medium Damage
DoD Criteria for
Antiterrorism
ASCE Blast
Design for
Petrochemical
Facilitiesa
UFC 3-340-02b
µMAX
θ MAX
µMAX
θ MAX
µMAX
θ MAX
Reinforced concrete beam
—
2
—
2
—
2c
Reinforced concrete slab
—
2
—
2
—
1
Reinforced masonry walls
—
2
—
2
Prestressed concrete
components
—
1d
—
1d
1
2
Element Type
1
Hot-rolled steel beam
3
3
10
6
10
2
Cold-formed steel beams
—
3
3
3
10
2
Corrugated steel panelse
—
1
3
2
6
4
Open-web steel joists
—
3
2
3
4
2
Wood studs and beams
2
—
—
—
—
—
Blast door components f
10–20
6
5–10
8
a
Response criteria from draft of new edition (to be published in 2009).
Design criteria for Protection Category 1. UFC 3-340-02 does not differentiate between cold-formed and hotrolled beams.
c
With shear reinforcement.
d
Reinforcement index ωp = (Aps /bd)(fps /f c ) < 0.15.
e
Panels attached to framing at both ends with some tension membrane resistance.
f
Door not operable but remains in opening.
b
3.5.7 Comparisons of Published Response Criteria
The response criteria discussed in this chapter are compared in this section. Cases
where different published response criteria are in agreement are indicative of
more accurate correlations between the given response levels and corresponding damage or response levels, in the sense that several different organizations
agree with those correlations. Cases where there is not agreement indicate that
more blast testing is needed for these component types, damage levels, and response modes to more definitively establish applicable response criteria. It must
also be considered that it is not always possible to have direct comparisons because of differences between the definitions of response levels by the different
organizations and the stipulations associated with various response criteria.
Table 3.18 shows comparisons of response criteria for Medium/Moderate
damage from the three sources discussed in this chapter. The criteria from UFC
3-340-02 are for design of Protection Category 1 (i.e., personnel protection),
which can be considered approximately equivalent to a medium damage level.
The comparison shows component types for which there is basic agreement of
the different criteria and component types for which there is not.
114
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
Table 3.19 Comparison of Response Criteria for High or Heavy Damage
DoD Criteria for
Antiterrorism
ASCE Blast
Design for
Petrochemical
Facilitiesa
UFC 3-340-02b
µMAX
θ MAX
µMAX
θ MAX
µMAX
θ MAX
Reinforced concrete beam
—
5
—
5
—
8c
Reinforced concrete slab
—
5
—
5
—
8
Reinforced masonry walls
—
8
—
5
Prestressed concrete
components
—
2d
—
2d
Hot-rolled steel beam
12
10
20
Cold-formed steel beams
—
10
12
Corrugated steel panels
—
4
Open-web steel joists
—
6
Wood studs and beams
3
Element Type
—
—
—
12
20
12
10
20
12
6
4
—
—
4
6
—
—
—
—
—
—
a
Response criteria from draft of new edition (to be published in 2009).
Design criteria for Protection Category 2. UFC 3-340-02 does not differentiate between cold-formed and hotrolled beams.
c
With shear reinforcement.
d
Reinforcement index ωp = (Aps /bd)(fps /f c ) < 0.15.
b
UFC 3-340-02 is generally more conservative than the other criteria, except
for corrugated steel panels. It is difficult to get a good comparison for this
case because the criteria in UFC 3-340-02 are for panels with “tension membrane capacity” without specifying any degree of tension membrane resistance,
as is implied in the other two criteria. The DoD and ASCE criteria agree with
each other fairly closely, except for hot-rolled steel beams and corrugated steel
panels.
Table 3.19 shows comparison of response criteria for High, or Heavy, damage from the three sources discussed in Section 3.5.1, 3.5.2, and 3.5.3.. The criteria from UFC 3-340-02 are for design of Protection Category 2 (i.e., equipment protection). All of the criteria in Table 3.19 are considered to represent
components that are at, or very near, complete failure where they would fall to
the ground or be thrown inside the building. In this case, the response criteria
in UFC 3-340-02 are not conservative compared to the criteria from the other
sources.
3.6 RESPONSE CRITERIA LIMITATIONS
A limitation of response criteria is that they are only applicable for very specific
response modes, since the amount of available ductility is a function of the mode
RESPONSE CRITERIA LIMITATIONS
115
that controls the component response for the specific case that is considered.
Unless otherwise stated, response criteria should be assumed to apply only for
components with ductile, flexural response where the component shear and connection capacities exceed maximum shear and end reaction forces, respectively,
based on the maximum resistance of the component during dynamic flexural response (i.e., equivalent static reaction forces).
This means that they are not applicable for describing damage from breach
(local failure through the full thickness in the component area of the highest blast
loads), spall (failure of the concrete near the surface in the component area of the
highest blast loads), or any other similar type of localized component response
mode. Breach and spall response are most commonly a concern for reinforced
concrete and masonry components subject to blast loads where the scaled standoff (i.e., charge standoff distance to the component divided by the charge weight
to the 1/3 power) is less than approximately 1.0 ft/lb1/3 . Several references, such
as Marchand et al. (Marchand et al. 1994), McVay (McVay 1988), and UFC
3-340-02 (UFC, 2008) have methods for predicting spall and breach in reinforced concrete slabs.
A component should be designed against blast loads so that ductile flexural
response controls the response, if possible, but if this is not possible, separate
response criteria that consider the specific type of response, such as response
controlled by shear, brittle flexure, combined flexure and axial load, or connection failure, as applicable, should be used. Unfortunately, such criteria are not
always available in published response criteria tables. Similarly, there are currently no published response criteria for many new materials that are used for
blast resistant design, such as fiber-reinforced plastic (FRP), ductile polymers,
and geotextile catch systems. In these situations, the response criteria must be
developed on a case-by-case basis by engineers with blast design and testing experience; otherwise, very inappropriate response criteria may result. For example, available response criteria for masonry walls reinforced with conventional
steel rebar have been proposed for walls reinforced with FRP, since these are
both cases of reinforced masonry, but this is very nonconservative because FRP
is much less ductile than reinforcing steel.
Another consideration is that the criteria in this chapter are intended primarily
for response calculated with SDOF analyses. However, they may also be used for
response calculated with more detailed dynamic analyses, such as multi-degreeof-freedom analysis and dynamic finite element analysis. Alternatively, design
based on dynamic finite element analysis can be based on more detailed response
criteria, such as maximum strain criteria, if the response criteria are validated
against adequate test data or are judged to be conservative enough by engineers
with significant applicable testing and blast design experience.
Finally, response criteria should also include any project-specific requirements, such as limits on deflection of a structural component from impacting
building equipment or other protected items. Another case in this category is design of blast-resistant shield walls, which are placed a few feet outside an existing
building wall to protect it from any blast loading. As the shield wall deflects very
116
PERFORMANCE CRITERIA FOR STRUCTURAL COMPONENTS
quickly from blast load, there is a volume decrease and corresponding pressure
buildup in the annular space between the shield wall and building. The resulting
pressure can cause the conventional building wall to fail if the deflection of the
shield wall is not adequately controlled. This is not considered in the available
response criteria.
The response criteria also do not address failure of any nonstructural items attached to blast-loaded structural components, where acceptable response of wall
and roof components per the applicable response criteria may cause large enough
relative accelerations between the component and attached item to rupture the
connection. This is most critical for heavy overhead items. This subject has not
been studied very much for blast-resistant design, but earthquake-resistant design
can be used for guidance.
REFERENCES
American Society of Civil Engineers. 1997a. Structural Design for Physical Security.
New York: American Society of Civil Engineers.
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997b. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
American Society of Civil Engineers Structural Engineering Institute. Forthcoming. Blast
Protection of Buildings. New York: American Society of Civil Engineers.
Applied Research Associates. 2008. Window Glazing Analysis Response and Design
(Wingard) Computer Program, Version 5.5.1. Developed by Applied Research Associates for the U.S. Government Services Administration (GSA).
ASTM. 2003a. Standard Practice for Specifying an Equivalent 3-Second Duration Design Loading for Blast Resistant Glazing Fabricated with Laminated Glass (F2248).
West Conshohocken, PA: ASTM International.
. 2003b. Standard Test Method for Metal Doors Used in Blast Resistant Applications (Equivalent Static Load Method) (F2247) West Conshohocken, PA: ASTM
International.
. 2007. Standard Practice for Determining Load Resistance of Glass in Buildings (ASTM E1300 - 07e1). West Conshohocken, PA: ASTM International.
Biggs, J. D. 1964. Introduction to Structural Dynamics. New York: McGraw Hill Co.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5-1300, Department of the Navy Publication NAVFAC P-397, Department of the
Air Force Manual AFM 88-22). Washington, DC: Departments of the Army, the Navy
and the Air Force.
Department of Defense Explosives Safety Board (DDESB). 2004. Ammunition and Explosives Safety Standards (DOD 6055.9). Washington, DC: U.S. Department of Defense.
Grumbach S. D., C. Naito, and R. J. Dinan. 2007. Use of precast concrete walls for blast
protection of steel stud wall construction. Presented at the 78th Shock and Vibration
Symposium, Philadelphia, PA, November 4–8, 2007.
Interagency Security Committee. 2004. ISC Security Design Criteria for New Federal
REFERENCES
117
Office Buildings and Major Modernization Projects. Washington, DC: Department of
Homeland Security, The Interagency Security Committee.
Marchand K. A., S. Woodson, and T. Knight. 1994. Revisiting concrete spall and breach
prediction curves: Strain rate (scale effect) and impulse (pulse length and charge shape)
considerations. Presented at the 26th DDESB Explosives Safety Seminar, Miami, FL,
August 1994.
McVay, Mark. 1988. Spall Damage of Concrete Structures. Vicksburg, MS: Army Engineer Waterways Experimentation Station Structures Lab. Accession No. ADA199225
from the Defense Technical Institute Center (DTIC).
Norville H. S. and E. J. Conrath. 2006. Blast-resistant glazing design. Journal of Architectural Engineering 12 (3): 129–136.
Oswald C. J. 2008. Update to response criteria for blast resistant buildings in petrochemical facilities. Presented at ASCE Structures Congress 2008, Vancouver, Canada, April
24–26, 2008.
Oswald C. J. and D. T. Nebuda. 2006. Development of Component Explosive Damage
Assessment Workbook (EDAW). Presented at the 32nd DDESB Explosive Safety Seminar, Philadelphia, PA, August 2006.
Oswald C. J., D. T. Nebuda, D. Holgado, and M. Diaz. 2006. Shock tube testing on
reinforced masonry walls. Presented at the 32nd DDESB Explosive Safety Seminar,
Philadelphia, PA, August 2006.
Panagiotakos T. B. and M. N. Fardis. 2001. Deformations of reinforced concrete members
at yielding and ultimate. ACI Structural Journal 98 (2): 135–148.
Protective Design Center. 2008a. Component Explosive Damage Assessment Workbook
(CEDAW) (PDC TR-08-07). Omaha, NE: U.S. Army Corps of Engineers, Protective
Design Center.
. 2008b. Single Degree of Freedom Structural Response Limits for Antiterrorism Design (PDC TR-06-08). Omaha, NE: U.S. Army Corps of Engineers, Protective
Design Center.
SAFE/SSG (Security Facilities Executive, Special Services Group). 1997. Glazing Hazard Guide, Cubicle Standoffs, Tables and Charts, Report No. SSG/EP/4/97. London:
Special Services Group/Explosive Protection Group.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (FOUO) (UFC 3-340-01). Washington, DC: U.S.
Department of Defense, Unified Facilities Criteria Program.
. 2005. Design of Buildings to Resist Progressive Collapse (UFC 4-023-03).
Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Army Corps of Engineers. 2004. Window Fragment Hazard Level Analysis (HazL),
Version 1.2. Vicksburg, MS: USACE Engineer Research and Development Center.
U.S. General Services Administration. 2003. Progressive Collapse Analysis and Design
Guidelines for New Federal Office Buildings and Major Modernization Projects (June
2003). Washington, DC: General Services Administration.
Wesevich J. W., C. J. Oswald, M. T. Edel, M. J. Lowak, and S. S. Alaoui. 2002. Compile and Enhance Blast Related CMU Wall Test Database (CMUDS). Prepared for the
Office of Special Technology by Wilfred Baker Engineering, Inc., Contract No. 4175600-C-0900, September 6, 2002.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
4
Materials Performance
Andrew Whittaker and John Abruzzo
4.1 INTRODUCTION
The ASCE Standard Blast Protection of Buildings (American Society of Civil
Engineers 2009) permits the use of reinforced concrete, masonry, structural steel,
and timber in blast-resistant construction. Of these materials, only the two most
common materials, namely, structural steel (Section 4.2) and reinforced concrete
(Section 4.3) are discussed in this chapter. For each of these materials, constitutive models that account for strain-rate and thermal effects, procedures to account
for strain-rate effects at the macro level, mechanical properties for design, and
failure modes at the component level are discussed. Only those grades of steel,
concrete and rebar materials, and components used in the construction of typical
buildings are discussed. Steel plates, light-gage and cold-formed framing are not
structural components in typical buildings, and the interested reader is directed
to UFC 3-340-02 (U.S. Department of Defense 2008) for relevant information.
Section 4.4 discusses the use of strength-reduction factors for the computation of strength of reinforced concrete and steel components and provides the
rationale for the use of φ = 1in the ASCE Standard.
A list of key references and resource documents is provided at the end of this
chapter.
4.2 STRUCTURAL STEEL
4.2.1 Stress-Strain Relationships
Generic uniaxial stress-strain relationships are available for structural steels in
textbooks (e.g., Bruneau et al. 1997, Brockenbrough and Merritt 2006) and on
the World Wide Web. Sample stress-strain relationships are presented in Figure
4.1 (from Brockenbrough and Merritt 2006). Such relationships are established
by ASTM-standard tests and can be described using yield stress (σ y ), tensile
(or ultimate) strength (σu ), yield strain (ε y ), strain at the onset of material hardening, strain at tensile strength, and strain at rupture. ASTM specifications for
structural steels (e.g., A36, A500, A572, A913, and A992) specify minimum
values for yield and tensile strengths and strain at rupture. Limits can be set on
119
120
MATERIALS PERFORMANCE
Figure 4.1 Stress-Strain Relationships for Selected Structural Steels at Room Temperature and Low Strain Rate (Brockenbrough and Merritt 2006, with permission)
maximum yield strength and maximum ratios of yield-to-tensile-strength can be
specified for selected ASTM-designated steel. Tests to determine stress-strain
relationships, σ = f (ε), are generally conducted at low speeds and produce results at which strain-rate effects are considered negligible. Data from such relationships, after modification for strain-rate effects, are routinely used to compute
mechanical properties for design.
Rapid or high strain-rate loading affects the mechanical properties of structural steels. High strain-rate data are reported in the literature (e.g., Campbell
and Ferguson 1970, Brockenbrough and Merritt 2006, Chang and Lee 1987,
Klepaczko 1994, Lee and Liu 2006). The typical effects of increased strain rate
on the response of structural steels are an increase in yield stress; an increase
in ultimate strength, albeit smaller than for yield stress; and a reduction in the
elongation at rupture. The elastic modulus is not substantially affected by strain
rate. Sample data for selected steels at three temperatures, from Brockenbrough
and Merritt (2006), are presented in Figure 4.2.
4.2.2 Constitutive Models for Structural Steel
Finite element analysis of steel components and structures for blast effects generally involves the use of either solid or shell elements to construct a beam (column) cross section and span or a beam-column element. If solid (shell) elements
STRUCTURAL STEEL
121
Figure 4.2 Effects of Strain Rate on Yield Stress and Tensile Strength at −50◦ F, Room
Temperature, and +600◦ F (Brockenbrough and Merritt 2006, with permission)
are used to discretize a cross section, as shown in Figure 4.3, an appropriate
constitutive model must be selected for the structural steel. If beam-column elements are used to describe a structural component, the effects of strain rate and
temperature must be accommodated indirectly, as described in Section 4.2.3.
Figure 4.3 Finite Element Model of a W-Shape Steel Cross Section
122
MATERIALS PERFORMANCE
400
293°K
LOWER YIELD STRESS τ (MNm–2)
195°K
300
225°K
IV
493°K
200
713°K
373°K
II
100
I
1
102
10
∝.
0
10–4 10–3 10–2 10–1
STRAIN RATE
(SEC
103
104
105
106
–1)
Figure 4.4 Effect of Strain Rate on Mild Steel (Campbell and Ferguson 1970, with
permission)
Most of the finite element codes used for commercial blast analysis of structural components and systems include a family of constitutive models for metals
that relate stress (σ ) and strain (ε), strain rate (˙ε ), and temperature (T), namely,
σ = f (ε, ε˙ , T )
(4.1)
Some of these models are empirical, with coefficients established by curvefitting to experimental data. Empirical models seek to reproduce test data collected at alternate strain rates and temperatures, such as the dataset of Figure 4.4
for mild steel reported by Campbell and Ferguson (1970). Others are physically
based. Meyer (1992) and Meyers (1994) identify many of the constitutive models
in use today.
One of the most widely used empirical models is that of Johnson and Cook
(1983) who proposed that the basic relationship between stress and strain at low
strain rate
σ = σ0 + Aεa
(4.2)
be modified by independent terms related to strain rate and temperature, namely,
σ = (σ0 + Aεa )(1 + BIn
ε˙
)(1 − [T ∗ ]b )
ε˙ 0
(4.3)
STRUCTURAL STEEL
123
where σ0 is the yield stress, a is a hardening coefficient, A and B are experimentally determined factors, ε˙ 0 is a reference strain rate, T ∗ is a normalized
temperature calculated per Equation 4.4, and b is a coefficient determined by
curve-fitting to data.
T∗ =
T − Tr
Tm − Tr
(4.4)
In Equation 4.4, T is the temperature at which the stress is calculated, Tr is a
reference temperature at which the yield stress is measured, and Tm is the melting
point. The five parameters (σ0 , A, B, a, b) in the model of Equation 4.3 are determined experimentally. Johnson and Cook (1983), Nicholas and Rajendran (1990)
and Meyers (1994) report values of these parameters for a number of materials,
but not structural grades of steel used in buildings, bridges, and infrastructure.
The Johnson-Cook model treats the effects of strain rate and temperature independently, which may be invalid for many metals at very high strain rates (Zukas
2004, Nicholas and Rajendran 1990).
Other constitutive models are physically based and provide a relationship between stress and strain that accounts for the effects of lattice structure, grain size,
strain rate, and temperature. Two models that are implemented in LS-DYNA
(Livermore Software Technology Corporation 2003) are Zerilli-Armstrong (Zerilli and Armstrong 1987, Lee and Liu 2006) and Mechanical Threshold Stress,
both of which are described by Meyers (1994) in some detail.
The use of either empirical or physically-based constitutive models in finite
element analysis enables the user to track the effects of strain rate and temperature over the depth and length of a cross section at each time step in the analysis.
4.2.3 Component Level Strain Rate and Temperature Effects
Most applications of the Standard will involve the use of simplified nonlinear
models of structural components and indirect treatment of strain-rate effects.
Thermal effects are generally ignored because (a) most computer codes used
to compute reflected pressure histories on components do not return temperature
histories, and (b) the time frame for the conduction of heat through the depth of
a steel component is often much longer than the duration of the imposed overpressure.
The most widely used nonlinear component model has a single-degree-offreedom (SDOF). The properties of the SDOF system are computed using procedures developed by Biggs (1964) and his co-workers at the Massachusetts Institute of Technology in the 1950s. These procedures can accommodate pinned
and fixed boundary conditions and typically assume a uniform distribution of
applied pressure, which is suitable for far-field detonations only. The response
of the nonlinear SDOF system is a function of the mass per unit length, flexural
stiffness, span, boundary conditions, history of reflected pressure, and component resistance function, which is a function of strain rate.
124
MATERIALS PERFORMANCE
Consider the cross section of Figure 4.3 and assume a linear distribution of
strain across the depth of the section. For W-shape cross sections subjected to
major-axis bending, axial strain and strain rate will be greatest at the exterior
edges of the flanges and zero at the neutral axis—strain rate will vary over the
depth of the cross section. For W-shape sections subjected to axial compression
or tension, axial strain and strain rate will be uniform across the depth of the
section.
UFC 3-340-02 (U.S. Department of Defense 2008) provides guidance on
the inclusion of strain-rate effects in the computation of component axial and
flexural strength in the form of stress-increase factors. Information is provided
for ASTM A36 (carbon structural steel), A588 (high-strength, low-alloy structural steel up to 345 MPa minimum yield point), and A514 (high-yield-strength
quenched and tempered alloy steel plate) steels, for flexure (bending) and axial
force (tension or compression) and low-pressure (far-range detonation, incident
overpressure less than 100 psi, lower strain rate) and high-pressure (near-range
detonation, incident overpressure much greater than 100 psi, higher strain rate)
loadings.
The UFC presents a simple equation for strain rate, which is assumed to be
constant from zero strain to the yield strain. The equation is reproduced in Equation 4.5, below. The assumption of constant strain rate to yield is not unreasonable given the coarseness of the analysis that generally accompanies the use of
the stress-increase factors.
ε˙ =
f ds
Ete
(4.5)
where ε˙ is the average strain rate in the elastic range of response, E is the elastic
modulus for steel (29,000 ksi, 200 GPa), te is the time to yield in the cross section, and f ds is the dynamic design stress, which is a function of minimum specified yield stress, a stress-increase factor, and a factor to relate expected strength
to minimum specified strength that varies as a function of steel grade. The dynamic design stress is discussed further in Section 4.2.4
Stress-increase factors, which are termed Dynamic Increase Factors (DIF) in
the UFC and denoted by the symbol c, are used to increase the yield stress and
tensile strength of structural steels. Table 4.1, which is adapted from UFC 3340-02, presents values for c. The values of c that are applied to yield stress are
larger for flexure than axial force because blast-induced axial forces are likely
applied indirectly through beam and girder reactions for which the rise time will
be longer and the load effects staggered in time. If a component is subjected to
combined axial force and bending moment (e.g., a column subjected to bending
due to direct air-blast loading on the column flange and axial force due to indirect
loading through air-blast-induced beam/girder reactions), one value of c should
be used for component strength calculations and deformation calculations, unless
the component yield strength can be updated step-by-step in the analysis. The
chosen value of c should reflect the dominant contributor to the combined stress
125
STRUCTURAL STEEL
Table 4.1 Dynamic Increase Factors for Structural Steel (Adapted from DoD
2008)
DIF, c, for Yield Stress
Specified
Minimum
Yield Stress1
(ksi)
Low
Pressure2
High
Pressure2
Low
Pressure3
High
Pressure3
DIF, c, for
Tensile
Strength
A36
36
1.29
1.36
1.19
1.24
1.10
A5884
50
1.19
1.24
1.12
1.15
1.05
A514
100
1.09
1.12
1.05
1.05
1.00
ASTM
Grade
Flexure
Axial
1
Typical value for standard W-shape sections.
Assumed strain rate of 0.10 s−1 and 0.30 s−1 for low and high pressure, respectively.
3
Assumed strain rate of 0.02 s−1 and 0.05 s−1 for low and high pressure, respectively.
4
The values presented in the shaded cells are estimates only.
2
ratio in the beam-column. The values of c for ultimate or tensile strength are
assumed to be independent of strain rate and are close to unity.
The dynamic increase factors presented in the last column of Table 4.1 are
close to unity and could be set equal to unity with no loss of robustness in the
resultant component design. Components subjected to blast loadings that sustain
strains associated with the tensile strength of the material likely do so near the
stage of maximum displacement, where the velocity and thus strain rate are close
to or equal to zero.
It is worthwhile to compare the strain-rate multipliers in UFC 3-340-02 of
Table 4.1 with data and equations provided by others. Moncarz and Krawinkler
(1981) provide an equation for the yield strength of ASTM A36 steel as a function of strain rate in the range of 0.0002 to 0.1 s−1 , namely:
σ y = σ0 (0.973 + 0.45˙ε0.33 )
(4.6)
where σ0 is the yield stress at a (low) strain rate of 0.0002, and all other terms
have been defined previously. Using this equation, the ratio of yield stress at ε˙ =
0.0002 to ε˙ = 0.1 s−1 is 1.2. For the data of Campbell and Ferguson of Figure
4.4, the ratio of yield stress at ε˙ = 0.0002 to ε˙ = 0.1 s−1 is also approximately
1.2 (at a temperature of 273◦ K or 20◦ C).
4.2.4 Mechanical Properties for Design
Moment-curvature relationships for steel cross sections have been developed using the data from uniaxial tension tests of steel coupons, but such relationships
generally ignore the presence of residual stresses in the cross section that result from the fabrication of the rolled section (Bruneau et al. 1997). Simplified
bilinear moment-curvature relationships are generally sufficient for blast analysis computations, as described below.
126
MATERIALS PERFORMANCE
The flexural, axial, and shear strengths of steel components are typically defined in terms of cross section properties (e.g., area, web area, elastic moduli,
plastic moduli) and either yield stress or tensile strength. For blast-resistant
design in accordance with the Standard, these stresses can be increased to a
dynamic design stress by factors to account for strain-rate effects (see Section 4.2.2) and material strengths in excess of minimum specified stresses or
strengths.
UFC 3-340-02 presents a widely accepted procedure for computing so-called
dynamic design stresses. The focus of the UFC presentation is flexure. Alternate computations are made for ductility ratios of less than or equal to 10 and
ductility ratios greater than 10. Therein, ductility is defined as the ratio of maximum deflection to yield deflection. Although large lateral deflections may be
acceptable to a designer, careful consideration must be given to geometry of
the cross section at these deflections, because the cross section may no longer
be prismatic due to either element deformation due to direct air-blast loading
(see Figure 4.5) or cross section instability caused by flange (web) local buckling or lateral-torsional buckling. If the component cross section is distorted, the
design equations presented herein, in UFC 3-340-02 and the AISC Steel Construction Manual (American Institute of Steel Construction 2006), cannot be
used for component checking.
For ductility ratios of 10 and less, the dynamic design stress, f ds , can be computed per Equation 4.7 as the product of the minimum specified yield stress, f y ,
a ratio of the expected yield stress to the minimum specified yield stress, R y , and
a Dynamic Increase Factor, c, namely,
f ds = f dy = f y R y c
(4.7)
where R y can be taken as either (a) 1.1 for f y of 50 ksi (345 MPa) or less and 1.0
otherwise, or (b) per Table I-4-1 of the AISC Seismic Provisions for Steel Buildings (American Institute of Steel Construction 2005) for new structural steel, c is
given in Table 4.1 or established on the basis of test data, and f dy is the dynamic
yield stress. For evaluation of older structural members constructed with A36
steel, R y should be set equal to 1.1 unless test data indicate that a greater value
is appropriate.
For ductility ratios of greater than 10, the UFC acknowledges the increase in
steel stress due to strain hardening per Equation 4.8. The relationship between
f ds and ductility ratio must be considered approximate because displacements
and peak strains do not scale linearly. Importantly, the technical basis for the
increase in stress beyond the dynamic yield stress is not provided in the UFC.
f ds = f dy +
f du − f dy
4
(4.8)
where f dy is the dynamic yield stress per Equation 4.7, and f du is the dynamic
ultimate stress. In the UFC, the dynamic ultimate stress is taken as the product
STRUCTURAL STEEL
127
Figure 4.5 Steel Cross Section Distorted Due to Air-Blast Loading (Courtesy of
American Institute of Steel Construction)
of the specified minimum tensile strength and c per Table 4.1. Alternately, the
dynamic design stress at high deformation can also account for the expected
increase in tensile strength above the specified minimum value, namely,
f ds = f dy +
c Rt f y − f dy
α
(4.9)
where values of R y are provided in Table I-4-1 of Seismic Provisions for Steel
Buildings (American Institute of Steel Construction 2005), and α is specified by
the user, but can be taken as 4.
4.2.5 Failure Modes of Structural Components
Introduction This section of the handbook presents introductory information
on typical failure modes (flexural response, shear response, combined flexural and axial response, instability) of structural shapes, with emphasis on
W-shapes loaded in the plane of the web. The focus is on design of new structural
components rather than assessment of existing components.
For assessment of existing structural components whose geometries and/or
materials do not comply with modern standards such as AISC (American
128
MATERIALS PERFORMANCE
Institute of Steel Construction 2005, 2006), the designer is encouraged to
use first-principles mechanics to determine whether performance is acceptable.
Knowledge of the component deformation demands may enable a relaxation of
the element compactness and component bracing rules required by Seismic Provisions for Structural Steel Buildings (AISC 2005) and adopted by-and-large
below.
Flexure Procedures for the computation of flexural strength of prismatic cross
sections as a function of unbraced length of the compression flange and element
compactness are well established and reported in textbooks and the literature. For
the simplest case, where the unbraced length is zero, the cross section is compact,
and the strain rate is low, the flexural strength is computed as the product of the
plastic section modulus and the minimum specified yield stress.
For component checking involving blast load effects, the flexural strength can
be calculated as the product of the plastic section modulus and the dynamic design stress per Section 4.2.3, provided the cross section is seismically compact
per Table I-8-1 of AISC (2005) and laterally braced per Section 9.8 of AISC
(2005). Lateral bracing of both flanges is generally required to address rebound.
Shear The capacity of a steel beam or column in shear can be computed as the
product of the shear area and the dynamic yield stress in shear. For a W-shape
loaded in the plane of its web, the shear area shall not exceed the web area. The
dynamic yield stress in shear, f dv , may be computed as
f dv = 0.60 f dy
(4.10)
where f dy is calculated per Equation 4.7. At the connection of such a W-shape
beam to the flange of a W-shape column, the distributions of stress and strain
will not follow beam theory, and a significant percentage of the shear force in
the beam will be transmitted to the column via the beam flanges (e.g., Kim et al.
2002).
Flexure and Axial Load For checking components subjected to blast load effects that produce flexure and axial load, the procedures of Section 9 of AISC
(2005) and AISC (2006) should be followed to compute coexisting axial and
flexural strengths.
Splices in components subjected to both flexure and axial load should develop
the capacity of the cross section computed assuming the dynamic design stress.
Flexure-axial-shear interaction need not be considered for a W-shape section
loaded in the plane of its web if the blast-induced shear force is less than shear
capacity of the cross section calculated using Equation 4.10 and the ratio of the
plastic section modulus of the web alone to that of the W-shape is small.
Element and Component Instability Element instability (e.g., flange and
web local buckling, web crippling and tearing) and/or component instability
REINFORCED CONCRETE
129
(e.g., lateral-torsional buckling) can prevent a cross section from attaining its
plastic capacity.
Design and detailing of new structural components to resist blast load effects should follow the AISC Seismic Provisions for Structural Steel Buildings
(AISC 2005) unless the component deformations are small. For component ductility demands of 3 and greater, cross sections should be seismically compact per
Table I-8-1 of the AISC Seismic Provisions. For component ductility demands
of less than 3, cross sections should be compact per Table B4.1of the AISC Steel
Construction Manual (AISC 2006).
Crippling or tearing of webs in W-shape cross sections may result from the application of concentrated loads by other framing members. Design against crippling or tearing should conform with Section J10 of AISC (2006), with no allowance for increase in yield stress above the expected value due to strain rate
because the rise time of the loading will likely be much longer than the duration
of the air-blast loading.
Lateral bracing of W-shape flanges to delay lateral-torsional buckling should
generally follow the rules of AISC (2005) for seismic design. Alternate limits
on unbraced length that vary as a function of the component deformation are
provided in UFC 3-340-02 and can be used in lieu of those presented in Section
9.8 of AISC (2005).
4.3 REINFORCED CONCRETE
4.3.1 Stress-Strain Relationships for Concrete
Many textbooks present stress-strain relationships for concrete in uniaxial compression. Sample relationships are presented in Figure 4.6 from Wight and
MacGregor (2008) for concrete with compressive strength ( f c ) ranging between
4500 psi (31 MPa) and 17500 psi (120 MPa). As with structural steel, the stressstrain relationships are established using an ASTM-standard test, which is conducted at a low stress rate of the order of 30 psi per second and causes failure
in 90 to 180 seconds. An increase in compressive strength is accompanied by
increase in the modulus of elasticity (secant modulus in the figure to a stress of
approximately0.5 f c ), an increase in the strain at maximum stress, an increase in
the slope of the softening branch of the stress-strain relationship, and a decrease
in the maximum concrete strain. The modulus of elasticity in units of psi for
normal-weight concrete with a density of 145 lb/ft3 is often taken per ACI 318
(American Concrete Institute 2008) as
E c = 57000 f c
(4.11)
where f c is in units of psi. Values of compressive strength are used to compute
mechanical properties for design after modification for strain-rate effects, as described in Section 4.3.5 below.
130
MATERIALS PERFORMANCE
Figure 4.6 Stress-Strain Relationships for Concrete in Uniaxial Compression at Low
Speed (From Wight and MacGregor 2008, with permission)
Wight and MacGregor present analytical approximations to the uniaxial
stress-strain relationship for concrete, which are not repeated here, and a stressstrain relationship for normal-weight concrete in tension, which is reproduced in
Figure 4.7. The tensile strength of concrete, f t , ranges between 8 and 15 percent
of the compressive strength, with the value dependant on the type of test used for
the measurement (Wight and MacGregor 2008).
The stress-strain relationship for plain concrete is a function of strain rate.
Bischoff and Perry (1991), Ross et al. (1995), Malvar and Ross (1998), Hentz
et al. (2004), Schuler et al. (2006), and Hao and Zhou (2007), among many others, have reported on the effect of strain rate on either compressive or tensile
strength. Sample results are presented in Figure 4.8 (Bischoff and Perry 1991)
for compressive strength and Figure 4.9 (Schuler et al. 2006) for tensile strength.
The vertical axis in each figure is the ratio of dynamic stress to static stress.
REINFORCED CONCRETE
131
Figure 4.7 Generic Stress-Strain Relationship for Concrete in Tension (From Wight
and MacGregor 2008, with permission)
Figure 4.8 Effect of Strain Rate on Compressive Strength (From Bischoff and Perry
1991, with permission)
132
MATERIALS PERFORMANCE
Figure 4.9 Effect of Strain Rate on Tensile Strength of Concrete (From Schuler et al.
2006, with permission)
Appreciable increases in both compressive and tensile strength are evident at
strain rates greater than 0.1 s−1 .
4.3.2 Stress-strain Relationships for Reinforcement
Typical stress-strain relationships for steel reinforcement (or rebar) are presented
in standards and textbooks. Figure 4.10 presents a typical stress-strain relationship for Grade 60 rebar from Malvar (1998). Grade 60 reinforcement is produced in accordance with a number of ASTM standards, including ASTM A615
and ASTM A704. ASTM A706 rebar is often used for seismic and blast applications where weldability, ductility, and bendability are important (Wight and
MacGregor 2008).
As with structural steel, the stress-strain relationship for steel reinforcement
is a function of strain rate. Sample data from Malvar (1998) showing the effect
of strain rate on yield stress and tensile strength are presented in Figure 4.11 and
4.12. The greatest increases in yield stress and tensile strength are observed for
the lower grades of rebar, and the percentage increases in yield stress are greater
than those in tensile strength.
4.3.3 Constitutive Modeling of Concrete and Rebar
Introduction Finite element models of reinforced concrete components and
structures for air-blast analysis will generally use either solid elements to mesh a
REINFORCED CONCRETE
133
Figure 4.10 Typical Stress-Strain Relationship for Grade 60 Steel Reinforcement
(From Malvar 1998, with permission)
cross section along a span or a beam-column element. If solid elements are used
to discretize a cross section, appropriate constitutive models must be selected
for the concrete and reinforcement, and algorithms and elements introduced to
model bond between the concrete and reinforcement. The following subsections
of Section 4.3.3 introduce constitutive models for plain concrete and reinforcement. If beam-column elements are used to model and analyze a reinforced
Figure 4.11 Effect of Strain Rate on Yield Stress of Steel Reinforcement (From Malvar
1998, with permission)
134
MATERIALS PERFORMANCE
Figure 4.12 Effect of Strain Rate on Tensile (Ultimate) Strength of Steel Reinforcement (From Malvar 1998, with permission)
concrete subjected to blast loading, the effects of strain rate (and temperature
if necessary) must be accommodated indirectly, as described in Section 4.3.4.
Plain Concrete Many material models for plain concrete at low strain rate have
been developed. Hao and Zhou (2007) note that these concrete material models
are often modified for high strain-rate loadings by multiplying important low
strain-rate properties such as compressive and tensile strength by a dynamic increase factor that is strain-rate-dependent. The interested reader is directed to
the literature, including Century Dynamics (2005), LSTC (Livermore Software
Technology Corporation 2003), Hao and Zhou (2007), and Tu and Lu (2009),
for information on these models, their implementation in hydrocodes and finite
element codes, and detailed bibliographies. Only increases in compressive and
tensile strength due to strain-rate effects are summarized below.
Recommendations for the compressive strength and/or tensile strength of concrete as a function of strain rate can be found in Soroushian et al. (1986), CEB
(Comit´e Euro-International du B´eton 1993), Malvar and Ross (1998), and Hao
and Zhou (2007), among others.
For compressive strength, CEB (1993) writes that the ratio of dynamic to
static strength (C(˙ε)), is
1.026α
ε˙
f cd
=
C(˙ε) =
f cs
ε˙ s
0.33
ε˙
=γ
ε˙ s
for ε˙ ≤ 30 s−1
(4.12)
for ε˙ > 30 s−1
REINFORCED CONCRETE
135
where f cd is the dynamic compressive strength at strain rate ε˙ in the range of
30 × 10−6 s−1 to 300 s−1 , f cs is the static compressive strength at a reference
strain rate ε˙ s of 30 × 10−6 s−1 , log γ = 6.156α − 2, α = 1/(5 + 9 f cs / f co ), and
f co = 1450 psi (=10 MPa). Malvar and Crawford (1998) noted that the CEB
model for compressive strength properly fitted the data available at the time of
their writing.
Gebbeken and Rupert (2000) proposed the multiplier of Equation 4.13 for
strength increases in compression (C) and tension (T),
Fm
− 1 Wy
C(η) = T (η) = (1 + tanh[0.4(log η − 2)]) ×
Wy
(4.13)
where η is a strain rate normalized by a reference value of 1 s−1 , log is the common logarithm, Fm is a variable that limits the increase at very high strain rates,
and W y is a variable that controls the shape of the function. Values of the variables are different in tension and compression (Gebbeken 2009).
Johnson and Holmquist (1992) presented the multiplier of Equation 4.14 for
strain-rate-related strength increases in tension and compression,
C(η) = T (η) = 1 + cLn(η)
(4.14)
where c is a constant equal to 0.007 (Hao and Zhou 2007), Ln is the natural
logarithm, and all terms have been defined previously. An identical equation
was adopted by Riedel et al. (1999) for the RHT model and implemented in
AUTODYN (Century Dynamics 2005).
For tensile strength, CEB (Comit´e Euro-International du B´eton 1990) writes
that the ratio of dynamic to static strength (T (˙ε )), is
1.016δ
ε˙
f td
=
T (˙ε ) =
f ts
ε˙ s
0.33
ε˙
=β
ε˙ s
for ε˙ ≤ 30 s−1
(4.15)
for ε˙ > 30 s
−1
where f td is the dynamic tensile strength at strain rate ε˙ in the range of 3 ×
10−6 s−1 to 300 s−1 , f ts is the static tensile strength at a reference strain rate ε˙ s of
3 × 10−6 s−1 , log β = 7.112δ − 2.33, δ = 1/(10 + 6 f cs / f co ), and f co = 10 MPa
(=1450 psi). Malvar and Ross (1998) modified the CEB equations as follows,
δ
f td
ε˙
=
for ε˙ ≤ 1 s−1
f ts
ε˙ s
0.33
ε˙
=β
for ε˙ > 1 s−1
ε˙ s
T (˙ε ) =
(4.16)
136
MATERIALS PERFORMANCE
where f td is the dynamic tensile strength at strain rate ε˙ in the range of 10−6 s−1
to 160 s−1 , f ts is the static tensile strength at a reference strain rate ε˙ s of 10−6 s−1 ,
log β = 6δ − 2, δ = 1/(1 + 8 f cs / f co ), and f co = 10 MPa (=1450 psi). Hao and
Zhou (2007) proposed an alternate formulation based on the data of Figure 4.9,
namely,
T (˙ε) = 1.0
= 2.06 + 0.26 log(˙ε)
= 2.06 + 2.0 log(˙ε)
for ε˙ ≤ 1 s−1
for 10−4 s−1 ≤ ε˙ ≤ 1 s−1
for 1 s−1 < ε˙ ≤ 103 s−1
(4.17)
where all terms are defined previously.
Rebar Materials models for rebar are generally identical to those for structural
steels as described in Section 4.2.2: Zhou et al. (2008) used the Johnson and
Cook (1983) material model to characterize the high-strain-rate response of rebar
in slabs constructed of conventional concrete and steel-fiber-reinforced concrete.
4.3.4 Component Level Strain-Rate Effects
Similar to steel structures, most reinforced concrete components will be analyzed and designed for air-blast effects using nonlinear single-degree-of-freedom
(SDOF) models and an indirect treatment of strain-rate effects, as introduced in
Section 4.2.3.
UFC 3-340-02 (U.S. Department of Defense 2008) provides guidance on the
inclusion of strain-rate effects for the computation of component axial, flexural,
and shear strength. The tensile strength of concrete is ignored for axial and flexural strength calculations. Strain-rate effects are addressed using stress-increase or
dynamic increase factors for material strength. Information is provided for concrete compressive and tensile strength and steel reinforcement yield stress and
tensile (ultimate) strength in alternate formats, including functions of strain-rate
and either far-design (low pressure) or close-in design (high-pressure) loadings.
As with calculations for structural steel, the UFC presents simple calculations
for average strain rate for concrete and rebar, which is assumed to be constant.
The calculations are based on an estimate of the time to yield of reinforcement
and the strain in the materials at yield of the cross section. For rebar, the average
strain rate is computed as
ε˙ s =
f dy
Ete
(4.18)
where ε˙ s is the average strain rate in the elastic range of response, E is the elastic
modulus for steel rebar (29,000 ksi, 200 GPa), te is the time to yield of the cross
section (generally computed using response charts), and f dy is the dynamic yield
stress, which is a function of minimum specified yield stress, a stress (dynamic)
REINFORCED CONCRETE
137
increase factor, and a factor to relate expected strength to minimum specified
strength that varies as a function of rebar grade. The variable f ds is the rebar
equivalent of the dynamic design stress for structural steel that was introduced
in Section 4.2.4. For concrete, the average strain rate, ε˙ c , is computed as
ε˙ c =
εc0
te
(4.19)
where εc0 is the strain at maximum concrete stress, which can be taken as 0.002
(see Figure 4.6).
Calculations of strain rate made using Equations 4.18 and 4.19 should be used
with care, in the knowledge that they are at best approximate. Consider a singly
reinforced concrete beam for which the neutral axis depth at yield of the cross
section is approximately 0.30 times the effective depth of the tension rebar, and
the yield strain in the rebar after accounting for strength in excess of the nominal
yield stress and rate effects is 0.0025: the maximum concrete strain at yield of
the tension rebar is approximately 0.0011, and the concrete strain at the approximate location of the resultant compressive force is approximately 0.0007, and
substantially smaller than the 0.002 recommended by the UFC. Alternate values
of strain rate for rebar and concrete would be calculated for doubly reinforced
cross sections (with differing values of strain rate in the tension and compression
rebar), columns subjected to blast loads producing flexure and compression or
tension, walls, and slabs.
Table 4.2, which is adapted from UFC 3-340-02, presents Dynamic Increase
Factors (DIF) for rebar (yield stress and tensile strength) and concrete in compression for flexure (bending), axial compression, shear (diagonal tension failure), shear (direct tension), and bond. The UFC writes that the values given in the
table for flexure assume a strain rate of 0.10 s−1 and 0.30 s−1 for both rebar and
concrete in the low-pressure (far-field) and high-pressure (near-field) ranges, respectively; for reinforced concrete members in compression (i.e., columns), the
Table 4.2 Dynamic Increase Factors for Reinforced Concrete (Adapted from DoD
2008)
Low Pressure
High Pressure
Rebar
Action
Yield
Rebar
Tensile
Concrete
Yield
Tensile
Concrete
Flexure
1.17
1.05
1.19
1.23
1.05
1.25
Compression
1.10
1.00
1.12
1.13
1.00
1.16
Shear—DT
1.00
1.00
1.00
1.10
1.00
1.00
Shear—Direct
1.10
1.00
1.10
1.10
1.00
1.10
Bond
1.17
1.05
1.00
1.23
1.05
1.00
138
MATERIALS PERFORMANCE
corresponding strain rates are 0.02 s−1 and 0.05 s−1 , respectively. These default
strain rates are identical to those assumed for structural steel and reported in the
footnotes to Table 4.1.
As with components of structural steel, the UFC recommends greater increases for rate effects for flexure than compression, noting that “slabs, beams
and girders filter the dynamic effects of the blast load . . . . Dynamic load reaching
columns . . . [has] . . . a long rise time of load . . . [which] results in lower strain
rate.” Further, the UFC writes that the values for shear and bond are smaller than
those for bending or compression for the purpose of “the need to prevent brittle
shear and bond failure and to account for uncertainties in the design process for
shear and bond.”
It is not clear whether the use of three significant figures in Table 4.2 is warranted. Given the range of strain rate associated with the categories of low and
high pressure, the use of two significant figures is likely appropriate. For lower
levels of protection, reinforced concrete components are permitted to sustain
maximum deformations associated with the tensile strength of the rebar. However, component velocity and thus strain rate at the stage of maximum deformation are zero, and material dynamic increase factors should be set to unity.
If a reinforced concrete component is subjected to combined axial force and
bending moment (e.g., a column subjected to bending due to direct air-blast loading on the column flange and axial force due to indirect loading through air-blastinduced beam/girder reactions), one value of the dynamic increase factor should
be used for component strength calculations and deformation calculations, unless the component yield surface can be updated step-by-step in the analysis.
The chosen value of the factor should reflect the dominant contributor to the
combined strength ratio in the beam-column.
4.3.5 Mechanical Properties for Design
The flexural, axial, and shear strength of reinforced concrete components are
typically computed using cross section properties and rebar stresses and concrete strengths. For blast-resistant design in accordance with the Standard, these
stresses and strengths can be increased to a dynamic design stress by factors to
account for strain-rate effects (see Section 4.3.3) and material strengths in excess
of minimum specified values.
UFC 3-340-02 presents procedures for computing dynamic design stresses
and strengths. As for structural steel, the stresses and strengths are tied to deformations but support rotation computed using SDOF analysis is used instead of
ductility ratio (see Section 4.2.4). Table 4.3 presents the UFC recommendations
for dynamic design stresses and strengths for the purpose of design, organized by
action: flexure, shear (diagonal tension and direct), and compression. For flexure and support rotations, θm , of less than 2 degrees, compression rebar can be
ignored for the calculation of flexural strength; for support rotations in excess of
2 degrees, the concrete is assumed to have crushed, and the flexural strength is
provided by a couple between the tension and compression rebar. An increase
REINFORCED CONCRETE
Table 4.3
2008)
Dynamic Design Stresses for Reinforced Concrete (Adapted from DoD
Action
Rebar Type
Flexure
Tension and
Compression
Support
Rotation,
Degrees
Rebar, f ds
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy 1
f dy + ( f du − f dy )/4
( f du + f dy )/2
Dynamic Design Stress
Shear—DT3
Stirrups
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
f dy
f dy
Shear—DT
Lacing
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
f dy + ( f du − f dy )/4
( f du + f dy )/2
Shear—Direct
Diagonal
0 < θm ≤ 2
2 < θm ≤ 6
6 < θm ≤ 12
f dy
( f du + f dy )/2
f dy
f dy
Compression
139
All
5
Concrete, f dc
f dc
2
2
f dc
f dc
f dc
f dc
f dc
f dc
f dc
4
4
f dc
1
For tension rebar only.
Concrete assumed crushed and moment resistance provided by tension-compression rebar couple.
3
Diagonal tension.
4
Concrete considered ineffective and shear resisted by diagonal rebar only.
5
Component capacity not a function of support rotation.
2
in support rotation corresponds to an increase in rebar strain (and stress), as
indicated in the table. The equation used to compute dynamic design stress for
support rotations 2 < θm ≤ 6 degrees is identical to Equation 4.8 for structural
steel at large deformations.
Dynamic design stresses (strength) for rebar are computed in a similar manner to that described in Section 4.2.4 for structural steel. Equation 4.7 is used
to compute the dynamic yield stress. For ASTM A615 Grade 60 and A706 rebar, the UFC recommends R y equal to 1.1, which results in an expected yield
stress of 66 ksi (455 MPa). Again, as with structural steel, the dynamic ultimate
stress for rebar is taken as the product of the specified minimum tensile strength
and a dynamic increase factor. For concrete, the dynamic compressive strength
is taken as the product of the specified (static) compressive strength and a dynamic increase factor; no allowance is made for concrete strength in excess of
the minimum specified value.
The dynamic increase factors for steel rebar and concrete can be established
using the values provided in Table 4.2, using the data presented in Sections 4.3.1
and 4.3.2, or using charts presented in the UFC, two of which are reproduced
below for concrete compressive strength (Figure 4.13) and rebar yield stress
(Figure 4.14). The curves provided in these figures express dynamic increase
140
MATERIALS PERFORMANCE
Figure 4.13 Effect of strain rate on compressive strength of concrete with f c in the
range of 2500 to 5000 psi (From DoD 2008)
Figure 4.14 Effect of strain rate on the yield stress and tensile strength of rebar (From
DoD 2008)
REINFORCED CONCRETE
141
factors as a function of strain rate. Figure 4.13 represents a reasonable lower
bound on the data of Bischoff and Perry and the CEB recommendations of Equation 4.12 for a concrete strength of 50 MPa (7250 psi) and strain rates of 10 s−1
and less. Note that the strain rate in this figure is presented in units of msec.
The curves of Figure 4.14 for the yield and tensile strength of rebar are in close
agreement with the recommendations of Malvar (1998).
4.3.6 Component-Level Failure Modes
Introduction This section of the handbook presents introductory information
on typical failure modes (flexural response, shear response, combined flexural
and axial response) of reinforced concrete members. The focus is on design of
new structural components rather than assessment of existing components. The
reader is referred to the literature for information on external reinforcement of
concrete elements for blast resistance using steel plate and carbon fiber polymeric materials (e.g., Crawford et al. 2001, Buchan and Chen 2007).
Exactly as with structural steel, the designer is encouraged to use firstprinciples mechanics to determine whether performance is acceptable. The failure mechanisms considered here assume that the core of the reinforced concrete cross section is intact. Spalling and scabbing may occur, and the depth
of reinforced sections should be restricted to core dimensions in such cases.
Reinforced concrete components are susceptible to brisance, or shattering, for
near-field detonations. The likelihood of shattering should be checked, and the
reader is directed to the literature for appropriate procedures.
The following subsections address the common failure modes of reinforced
concrete components under blast loading. Much additional information on
failure modes can be found in the literature, including UFC 3-340-02 (U.S.
Department of Defense 2008), Krauthammer’s Modern Protective Structures
(Krauthammer 2008), Mays and Smith’s Blast Effects on Buildings (Mays and
Smith 1995), and Smith and Hetherington’s Blast and Ballistic Loading of Structures (Smith and Hetherington 1994).
Flexure Procedures for the computation of flexural strength of prismatic cross
sections are well established. Flexural failure can occur in either brittle or ductile modes, depending upon the volume of steel reinforcement in the flexural
element. Economical design of reinforced concrete components to resist blast
loads generally requires ductile detailing to accommodate the large expected deformations while maintaining near-peak strength. Confinement in the form of
seismic hoops or lacing may be required.
For component checking involving blast effects, the flexural strength can be
computed using the dynamic design stresses of Table 4.3 and compressiontension force couples. If the support rotations are less than 2 degrees, flexural
strength is computed using traditional procedures that assume that the tension
rebar yields prior to the concrete in the compression block crushing. For support rotations in the range of 2 to 6 degrees, the cover concrete is assumed to be
142
MATERIALS PERFORMANCE
lost, equal amounts of compression and tension rebar should be provided, and
the flexural strength is computed using a force-couple involving the compression and tension rebar. For support rotations in excess of 6 degrees and less than
12 degrees, equal amounts of compression and tension rebar must be provided,
the flexural strength is computed as a force-couple between the compression and
tension rebar, and the rebar layers must be adequately tied. Shear reinforcement
in the form of ties (stirrups) or lacing is generally required to sustain support
rotations greater than 2 degrees. UFC 3-340-02 provides guidance on the design
of such reinforcement.
For robustness, the minimum longitudinal reinforcement in slabs and beams
must meet the requirements of ACI 318 for load combinations not involving
blast effects. UFC 3-340-02 also requires a minimum amount of tension and
compression flexural rebar to be provided in slabs, but not less than that required
to resist the effects of direct air-blast loading and rebound. These limits for slabs
tend to be smaller than those of Section 10.5 of ACI 318, although a minimum
amount of compression rebar must be provided in all cases.
Shear Four modes of shear failure in reinforced concrete components must be
addressed: (1) diagonal tension, (2) diagonal compression, (3) direct shear, and
(4) punching shear. Critical cross sections are at a distance (a) equal to the effective depth of the cross section from the support for diagonal tension and compression checks for loadings that produce compression in the support and at the
face of the support otherwise, and (b) at the face of the support for direct shear.
Punching shear is checked for slabs around isolated supports (e.g., columns) and
applied loads. Concrete and rebar strengths for component checking can be increased for strain-rate effects per Tables 4.2 and 4.3.
Design against failure by diagonal compression or tension follows the procedure set forth in Chapter 11 of ACI 318. Diagonal compression failure is avoided
in beams (slabs) equipped
stirrups by limiting the nominal shear stress at
with
, where f dc
is the concrete dynamic design stress
the critical section to 10 f dc
per Tables 4.2 and 4.3. Diagonal tension failure is avoided by ensuring that the
sum of the shear resistance provided by the concrete and the stirrups exceeds applied shear force. ACI 318 Sections 11.2.1.1, 11.2.1.2, and 11.2.2.3.can be used
to compute the shear resistance provided by concrete under no axial load, axial compression, and axial tension, respectively. ACI 318 Sections 11.4.7.2 and
11.4.7.4 present equations for establishing the required area of shear reinforcement perpendicular to the axis of the member (i.e., stirrups) and inclined to the
axis of the member (i.e., lacing bars), respectively. The effective depth of the section used in these calculations of concrete and rebar shear resistance is a function
of the expected support rotation and the likelihood of loss of cover concrete. Section 4-18.2 of UFC 3-3340-02 provides appropriate guidance for the calculation
of effective depth.
UFC 3-340-02 requires that shear reinforcement be provided in slabs that
are required to develop large deflections (associated with support rotations
greater than 2 degrees). Shear reinforcement must always be provided to resist
REINFORCED CONCRETE
143
shear stresses in excess of the shear capacity of the concrete. Section 4.18.4
of the UFC provides appropriate guidance for minimum shear reinforcement
in slabs.
Direct shear failure has been observed in blast testing of reinforced concrete
slabs. Failure occurs along a vertical plane at load or stiffness discontinuities
such as the intersection of a slab and its supporting beam or wall. Krauthammer
(2008) provides a discussion of direct shear failure and presents a model to predict the relationship between imposed shear force and deformation (slip across
a vertical plane). Section 4.19 of UFC 3-340-02 presents design equations for
direct shear capacity that includes contributions from concrete and diagonal (inclined) bars. (Stirrups that are placed perpendicular to the plane of a slab provide
no resistance to direct shear because the failure plane is vertical.) Concrete is
assumed to provide resistance to direct shear per Equation 4.20 if the support
rotations are less than 2 degrees (i.e., there is little to no damage to the cover
concrete) and the slab is not subjected to net tension:
bd
Vc = 0.16 f dc
(4.20)
where b is the slab width and d is the effective depth of the cross section. For
support rotations greater than 2 degrees, the direct shear must be resisted by
diagonal (inclined) reinforcing bars only, where the required rebar area per width
b is
Ad =
(Vu − Vc )
f ds sin α
(4.21)
where Ad is the area of inclined rebar per width b, α is the angle between the
inclined rebar and the plane of the longitudinal rebar, Vu is the direct shear force
per width b, Vc is the concrete resistance width b, and f ds is the dynamic design stress for the inclined rebar per Table 4.3. The UFC sets no upper bound
on the applied direct shear stress, likely because the direct shear force is re
. Further, the
sisted by inclined bars only for shear stress in excess of 0.16 f dc
UFC equations do not recognize the benefit of axial compression on the cross
section, which will be present in some components (e.g., columns) and not in
others (e.g., slabs). Upper bounds on direct shear stress are identified in the literature with sample values of 0.35 f c (Ross and Krawinkler 1985) for static loads
(Crawford et al. 2001) for blast loadings. Normal compressive force
and 0.6 f dc
could be included in the formulation using an approach similar to the shearfriction design method of Section 11.6 of ACI 318 or as proposed by Crawford
et al. (2001).
Punching shear must be checked for slabs supported on isolated supports
such as columns. The procedures of Chapter 11 of ACI 318 should be used
to design components to avoid punching shear failure. Section 4.20.1 of
UFC 3-340-02 provides appropriate guidance for the calculation of effective
depth for the punching shear check.
144
MATERIALS PERFORMANCE
Flexure and Axial Load Reinforced concrete components subjected to flexure
and axial load can fail due to either excessive concrete strain (a compressiondominated component) or rebar rupture (flexure-dominated component). The latter can be treated as a flexural element with consideration of axial load. The
procedures of Chapter 10 of ACI 318 should be used to proportion axial-loaddominated components. Appropriate consideration should be given to strain-rate
effects per Section 4.3.4 to compute material stresses for component strength
calculations.
Reinforced concrete components such as slabs may be restrained against
movement in their plane by adjacent framing. Such restraint can give rise
to membrane effects. Krauthammer (2008) notes that compressive membrane
forces at relatively low lateral displacements can enhance the ultimate capacity
of a slab by increasing the flexural strength of the slab at the critical sections.
At large lateral displacements, associated with the loss of core concrete, resistance to blast loadings can be provided by the layers of rebar acting as a cable
net. For such a net to form, the rebar must be appropriately developed into the
supports, and the adjacent framing must be capable of developing the associated
reactions.
Scabbing Scabbing is the loss or separation of cover concrete from the core
concrete associated with large deformations in the component.
Spalling Spalling is similar to scabbing but occurs as a result of the reflection
of a compressive shock wave from the rear face of a concrete component. If the
resulting tensile stresses exceed the tensile strength of the concrete, concrete will
spall. The likelihood of spalling can be computed using codes such as CONWEP
(Hyde 1992).
4.4 STRENGTH-REDUCTION FACTORS FOR STEEL
AND REINFORCED CONCRETE
Strength-reduction factors, denoted φ in most materials standards, are used to
pre-multiply nominal component strength to compute design strength. Design
strengths are generally compared with factored load effects to determine the adequacy of a design. The values assigned to strength-reduction factors range between 0.6 and 0.9, with smaller values assigned to component actions that are
brittle (e.g., shear) and larger values to component actions that are ductile (e.g.,
flexure in the absence of axial force).
There is no industry-wide consensus on the values of φ for blast-resistant
design. Traditional practice has set φ equal to 1.0, and this decision can be
rationalized on the bases that (a) the weapon effect (air-blast) is idealized and
likely conservative, especially for near-field detonations, (b) equivalent SDOF
REFERENCES
145
models are used for load-effect calculations, and (c) strain-rate effects are estimated conservatively.
Blast analysis calculations performed using a finite element code use material
constitutive models to assess stress and strain at the cross section level, deformations at the component level, and displacements at the global level. Best estimate
(mean) material properties are used as input to these constitutive models with no
reduction to design values, that is, φ = 1.
REFERENCES
American Concrete Institute (ACI). 2008. Building Code Requirements for Structural
Concrete (ACI-318-08) and Commentary (ACI 318R-08). Farmington Hills, MI:
American Concrete Institute.
American Institute of Steel Construction (AISC). 2005. Seismic Provisions for Structural Steel Buildings (ANSI/AISC-341s1-05). Chicago, IL: American Institute of Steel
Construction.
. 2006. Steel Construction Manual. 13th Ed. Chicago, IL: American Institute of
Steel Construction.
. 2009. Blast Protection of Buildings (ASCE Standard). Reston, VA: American
Society of Civil Engineers.
Biggs, J. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill.
Bischoff, P. H. and S. H. Perry. 1991. Compressive behavior of concrete at high strain
rates. Materials and Structures 24 (6): 425–450.
Brockenbrough, R. L. and F. S. Merritt. 2006. Structural Steel Designers Handbook. New
York: McGraw-Hill.
Bruneau, M., C. M. Uang, and A. S. Whittaker. 1997. Ductile Design of Steel Structures.
New York: McGraw-Hill.
Buchan, P. A. and J. F. Chen. 2007. Blast resistance of FRP composites and polymer
strengthened concrete and masonry structures—A state-of-the-art review. Composites
Part B: Engineering 38 (5–6): 509–522.
Campbell, J. D. and W. G. Ferguson. 1970. The temperature and strain rate dependence
of the shear strength of mild steel. Philosophical Magazine 21 (169): 63–82.
Century Dynamics. 2005. AUTODYN Users Manual Version 6.1. Horsham, West Sussex,
UK: Century Dynamics.
Chang, K. C. and G. C. Lee. 1987. Strain rate effect on structural steel under cyclic
loading. Journal of Engineering Mechanics 113(9): 1292–1301.
Comit´e Euro-International du B´eton (CEB). 1993. CEB-FIP Model Code 1990. Trowbridge, Wiltshire, UK: Redwood Books.
Crawford, J. E., L. J. Malvar, K. B. Morrill, and J. M. Ferritto. 2001. Composite retrofits
to increase the blast resistance of reinforced concrete buildings (P-01-13). Proceedings
of the 10th International Symposium on Interaction of the Effects of Munitions with
Structures 2001. San Diego, CA.
Gebbeken, N. and M. Ruppert. 2000. A new material model for concrete in high dynamic
hydrocode simulations. Archive of Applied Mechanics 70 (7): 2509–2524.
146
MATERIALS PERFORMANCE
Gebbeken, N. 2009. Personal communication.
Hao, H. and X. Q. Zhou. 2007. Concrete material model for high-rate dynamic analysis. Proceedings of the 7th International Conference on Shock and Impact Loads on
Structures, Beijing, October 2007.
Hentz, S. F., V. Donze, and L. Daudeville. 2004. Discrete element modelling of concrete
submitted to dynamic loading at high strain rates. Computers and Structures 82 (29 30):
2509–2524.
Hyde, D. W. 1992. User’s Guide for Microcomputer Program CONWEP, Application of
TM5-855-1: Fundamental of Protective Design for Conventional Weapons. Instruction
Report SL-88-1. Vicksburg, MS: Structures Laboratory, U.S. Army Waterways Experiment Station.
Johnson, G. R. and W. H. Cook 1983. A constitutive model and data for metals subjected to large strains, high strain rates and high temperatures” Proceedings of the 7th
International Symposium on Ballistics 1983. The Hague, Netherlands.
Johnson, G. R. and T. J. Holmquist. 1992. A computational constitutive model for brittle
materials subjected to large strains, high strain rates and high pressures. In Shock-Wave
and High-Strain-Rate Phenomena in Materials, 1075–1081, edited by M. A. Meyers,
L. E. Murr, and K. P. Staudhammer. New York: Dekker.
Kim, T., A. S. Whittaker, A. J. S. Gilani, V. V. Bertero, and S. M. Takhirov. 2002. Coverplate and flange-plate steel moment-resisting connections. Journal of Structural Engineering 128 (4): 483–491.
Klepaczko, J. R. 1994. An experimental technique for shear testing at high and very
high strain rates. The case of a mild steel. International Journal of Impact Engineering
15 (1): 25–39.
Krauthammer, T. 2008. Modern Protective Structures. Boca Raton, FL: CRC Press.
Lee, W.-S. and C.-Y. Liu. 2004. The effects of temperature and strain rate on the dynamic
flow behavior of different steels. Materials Science and Engineering A 426 (1–2):
101–113.
Livermore Software Technology Corporation (LSTC). 2003. LS-DYNA User’s Manual.
Livermore, CA: Livermore Software Technology Corporation.
Malvar, L. J. 1998. Review of static and dynamic properties of steel reinforcing bars. ACI
Materials Journal 95 (5): 609–616.
Malvar, L. J. and J. E. Crawford. 1998. Dynamic increase factors for concrete. Proceedings of the 28th Department of Defense Explosives Safety Seminar 1998. Orlando,
FL.
Malvar, L. J. and C. A. Ross. 1998. Review of strain rate effects for concrete in tension.
ACI Materials Journal 95 (6): 735–739.
Mays, G. C. and P. D. Smith. 1995. Blast Effects on Buildings. London: Telford.
Meyer, L. W. 1992. Constitutive equations at high strain rates. In Shock-Wave and HighStrain-Rate Phenomena in Materials, 49–68, edited by M. A. Meyers, L. E. Murr, and
K. P. Staudhammer. New York: Dekker.
Meyers, M. A. 1994. Dynamic Behavior of Materials. New York: Wiley.
Moncarz, P. and H. Krawinkler. 1981. Theory and application of experimental model
analysis in earthquake engineering (Report No. 50). Palo Alto, CA: John A. Blume
Earthquake Research Center, Stanford University.
REFERENCES
147
Nicholas, T. and A. M. Rajendran. 1990. Material characterization at high strain rates.
Chapter 3 in High Velocity Impact Dynamics, edited by J. Zukas. New York: WileyInterscience.
Reidel, W., K. Thoma, S. Hiermaier, and E. Schmolinske. 1999. Penetrating of reinforced
concrete by BETA-B-500 numerical analysis using a new macroscopic concrete model
for hydrocodes. Proceedings of the 9th International Symposium on the Interaction of
the Effects of Munitions with Structures 1992, 315–322, Berlin.
Ross, C. A., J. W. Tedesco, and S. T. Kuennen. 1995. Effects of strain rate on concrete
strength. ACI Materials Journal 92 (1): 37–47.
Ross, T. J. and H. Krawinkler. 1985. Impulsive direct shear failure in RC slabs. Journal
of Structural Engineering 111 (8): 1661–1676.
Schuler, H., C. Mayrhofer, and K. Thoma. 2006. Spall experiments for the measurement
of the tensile strength and fracture energy of concrete at high strain rates. International
Journal of Impact Engineering 32 (10): 1635–1650.
Smith, P. D. and J. G. Hetherington. 1994. Blast and Ballistic Loading of Structures.
Oxford: Butterworth-Heinemann.
Soroushian, P., K. B. Choi, and A. Alhamad. 1986. Dynamic constitutive behavior of
concrete. ACI Journal 83 (2): 251–259.
Tu, Z. and Y. Lu. 2009. Evaluation of typical concrete material models used in hydrocodes
for high dynamic response simulations. International Journal of Impact Engineering
36 (1): 132–146.
U.S. Department of Defense (DoD). 2008. Structures to Resist the Effects of Accidental Explosions (United Facilities Criteria UFC 3-340-02). Washington, D.C.: U.S.
Department of Defense.
Wight, J. K. and J. G. MacGregor. 2008. Reinforced Concrete: Mechanics and Design.
5th Ed. Upper Saddle River, New Jersey: Prentice Hall.
Zerilli, F. J. and R. W. Armstrong. 1987. Dislocation-mechanics-based constitutive
relations for material dynamics calculations. Journal of Applied Physics 61 (5):
1816–1825.
Zhou, X. Q., V. A Kuznetsov, H. Hao, and J. Waschl. 2008. Numerical prediction of
concrete slab response to blast loading. International Journal of Impact Engineering
35 (10): 1186–1200.
Zukas, J. A. 2004. Introduction to Hydrocodes. Oxford: Elsevier.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
5
Performance Verification
Curt Betts
5.1 INTRODUCTION
The purpose of this chapter is to familiarize designers who are interested in blast
design with verification of performance for blast and weapons effects and for
stopping vehicles using barriers. The purpose is not to provide guidance on how
to do any of that testing. It is merely to make designers understand when verification and validation should be provided, to make them effective consumers of
performance verification services, and to ensure they understand how to validate
claims made either by other designers or by vendors. This chapter will address
verification by both testing and analysis. It will also address the topic of peer
review.
5.2 PERFORMANCE VERIFICATION
Before the subject of performance verification can be addressed, it is important to
point out that verification has a specific meaning in common use, and it is different from validation. Both involve confirmation, but by convention they confirm
different things in different ways. While their differences in meaning have been
the subject of controversy for decades, the following convention will be used in
this chapter.
Verification refers to the practice of ensuring that a model works for a given
set of conditions. For example, in blast-resistant designs, members may be assumed to be loaded uniformly within a particular scaled range, and they may be
assumed to respond to those loadings in flexure. If a member is closer than that
scaled range, models based on uniform loadings or flexural responses are not
applicable for the member response. Those models cannot be verified for that
application. Verification can also apply to ensuring that a given configuration
falls within the assumptions for meeting a particular performance condition. For
example, for tensile membrane action to be applicable, certain detailing needs
to be verified. Yet another way of applying verification is by checking that all
design requirements have been addressed, that calculations are correct, and that
the design actually reflects what the design analysis says it should.
149
150
PERFORMANCE VERIFICATION
Validation refers to the practice of ensuring that models are technically correct
by comparing them to other validated models or test results. For example, there
are many validated single-degree-of-freedom models for use in blast-resistant
design. Any new model should be validated against previously validated models by running identical member configurations and loadings to ensure that both
codes give similar answers. Alternatively, code results can be compared against
test data to ensure that the models give answers that are consistent with the test
data. Of course, that assumes that the test data are themselves validated or verified, which will lead to the next section of this chapter.
5.3 TESTING
This section will address testing of both building components and vehicle barriers. In both cases there are validated test procedures that can be used to evaluate
performance. Those test procedures are commonly based on some sort of standard test, such as one included in a national standard.
5.3.1 Vehicle Barrier Testing
Vehicle barriers are discussed in Chapter 12, where they are described as antiram structures. Due to the complexities of the structural systems associated with
such barriers, the soil–structure interaction, and the vehicle–structure interaction, designing vehicle barriers solely through analytical means is not sufficient
for validation. Analytical means can be used to design the barriers, but proper
validation requires those models to be validated by vehicle testing.
Any validation testing of vehicle barriers must be based on a standard test
method to be repeatable, to avoid bias, and to ensure that common performance
measures are reported for all vehicle barriers. One such test standard is ASTM
F2656, Vehicle Crash Testing of Perimeter Barriers. It, like any good standard
test, includes specific requirements on how the tests must be set up, how the tests
are run, and how performance is measured. Unlike some standard test methods,
there is a degree of subjectivity to vehicle barrier tests, which is controlled by the
test director who has to decide such things as where the barrier is impacted and
then to determine the validity of the test. Issues such as those are why such standards commonly require testing laboratories to be accredited. Figure 5.1 shows
the end result of a vehicle barrier test.
ASTM F2656, as an example of a test standard, includes details on the vehicles themselves, including gross vehicle weight, age of the vehicles, vehicle
soundness, how ballast needs to be assembled and attached, and in the case of
one of the vehicles, the wheel base. The test standard further establishes requirements for the information to be provided to the test director on the barrier and
information on how the barrier needs to be installed. The standard further establishes detailed criteria for measuring the performance of the barriers, including
the information that has to be collected and evaluated by the test director. That
TESTING
151
Figure 5.1 Vehicle Barrier Test
information includes vehicle acceleration, velocity, penetration, and vehicle and
barrier deformation. It also includes the apparatus necessary to measure those
quantities.
While test standards like ASTM F2656 provide for validation of designs, there
is also an element of verification associated with the barriers. Verification comes
into play when a barrier is specified for installation after testing. Verification at
that point includes reviewing the barrier drawings to ensure that the barrier that
is being installed is the barrier that was tested, that it is installed in a manner that
is consistent with the tested configuration, and that its deployment will result in
its being employed in a manner that is consistent with the way it was tested.
5.3.2 Building Components
Testing of building components similarly requires standard testing procedures to
ensure that tests are run consistently and with standard, comparable performance
measures. Tests can be run at full scale or sub-scale, and using either actual
explosives or blast simulators.
Full-scale testing is the most accurate method to test building components.
It involves building a component to be tested or an entire building to the actual
scale with the threat explosive, and at the applicable standoff distance. Through
152
PERFORMANCE VERIFICATION
Figure 5.2 Full–Scale Building Testing
such tests the full-scale behavior of the building or building component can be
measured or observed to validate compliance with a performance standard.
One common means to set up such a test is to build a building or a test frame
in which the articles to be tested are installed, and to place the threat explosive
at the standoff distance at which the response is to be observed. In the case of
building a whole building, its response and the response of its components are
directly and unambiguously observed. Figure 5.2 shows a full-scale building that
was tested.
In the case of a test frame, the results may need to be evaluated carefully.
Where such a test frame is narrow in comparison to how it will be installed in a
real building, a phenomenon referred to as “clearing” may result in inaccuracies
in the observed component response. Clearing may be an issue when reflected
pressure effects are being measured. With small or narrow targets, the reflected
pressure may be created on the front face of the target, but it may not persist
because the finite boundaries of the target allow the blast wave to propagate
around its edges such that the duration of the load in the target is reduced (the
clearing time), resulting in a reduced load on the component compared to what
it would see were it in a much wider or infinite surface. Clearing is not always
an issue, however. For components such as windows that commonly respond
very quickly to blast waves, the clearing time may not be an issue. Because of
the potential uncertainties in the component response, results from tests where
clearing is an issue need to be evaluated carefully. Figure 5.3 shows test frames.
TESTING
153
Figure 5.3 Test Frames in Arena Configuration
Full-scale testing is often done in what is referred to as an arena test, as shown
in Figure 5.3. In such a test, a number of test objects or test cells or frames
are placed around a test explosive in roughly a circular pattern. Different test
specimens may be at different standoff distances from the explosive, allowing
multiple results from the same test.
Sub-scale testing involves reducing the scale of dimensions associated with
the test, including the test specimen, the distance from the explosive to the test
specimen, and the explosive itself. In those tests, dynamic similitude must be
maintained, which for explosives is done according to what is called Hopkinson
scaling. In Hopkinson scaling, quantities are scaled in proportion to the cube root
of the explosive weight. By scaling, similar blast phenomena can be observed
with smaller explosives at closer distances. For example, refer to Table 5.1 below,
which shows the pressures and impulses for two different explosive weights at
two different standoff distances.
Note that the scaled range (standoff distance / the cube root of the explosive weight) is the same for both tests and that the pressures are the same. That
shows that a test can be run for one explosive weight and can then be scaled upward or downward to reflect the effects of different explosive weights at different
distances. This is a common practice in the blast community. Similarly, other dimensions such as wall heights can be scaled, allowing a wide range of options in
scaled modeling and testing.
It is important to note, however, that not all quantities scale alike. Note that
the impulses associated with the two equivalent pressures are not the same. This
illustrates why scaled testing has to be evaluated and verified carefully. While
such a pair of tests as those in Table 5.1 may be valid for components that are
Table 5.1 Hopkinson Scaled Pressures and Impulses
Quantity
Test #1
Test #2
Weight (W)
1000 lbs TNT
220 lbs TNT
Range (standoff distance)
100 ft
60.4 ft
Scaled range (Range / W1/3 )
10
10
W 1/3
10 lb 1/3
6.04 lb 1/3
Pressure
9.56 psi
9.56 psi
Impulse
80.9 psi-msec
48.7 psi-msec
154
PERFORMANCE VERIFICATION
most responsive to peak pressures, they may not be valid for components that are
more responsive to impulse.
In either full-scale or sub-scale testing, the phenomena that are to be observed
need to be measured carefully and accurately. Most importantly, it is imperative
that the measurements that best reflect the response to be observed are measured.
In all cases the basic geometry of the test needs to be recorded accurately, which
includes the dimensions and construction of the test subject, how it installed into
the test cell or target building, the construction of the test cell or building, the
size and composition of the explosive, and the distance and angle of incidence
from the explosive to the target. In addition, atmospheric conditions such as temperature, humidity, altitude, and wind speed and direction must be recorded.
In measuring test results, there are many measurements that a test report
should include, depending on the nature of the test. Very importantly, the blast
wave must be fully characterized using pressure gages and similar measuring
devices. Then the test specimen response parameters must be measured. Those
measurements depend on the response to be observed. If the member is intended
to remain intact with either elastic or inelastic deformation, the deflection of the
member must be measured. That will allow for calculation of support rotations
or ductility ratios (see Chapter 3). Those are commonly measured using strain
gages or displacement transducers.
If the response is expected to include failure of the test specimen with some
expected hazard level of the resulting fragments, additional measurements would
be required. Those would include measuring the sizes, weights, and translations
of the fragments, and their accelerations. Measuring devices for those include
accelerometers and high-speed digital cameras. Figure 5.4 shows a setup for both
high-speed digital camera recording and deflection. The camera cannot be seen,
but the stripes painted in the wall allow for accurate evaluation of the velocities
of the fragments.
Another option for measuring blast response of components is to use a blast
simulator. The most common blast simulators are tubes into which test specimens are mounted at one end. The blast simulating “driver” is at the end of the
tube opposite the test specimen, and it commonly drives air into the test specimen
at high velocities that simulate a blast wave. The measurements of the responses
of the test specimens are similar to those in open-air tests. Tests in blast simulators can usually be done much less expensively and in much quicker progression
than open-air blast tests, and they can be operated in areas with far fewer restrictions than those in which open-air blasts tests are run. It is important to recognize,
however, that the blast simulator loadings should be validated against open-air
tests to ensure that the loadings are representative of actual conditions. Figure
5.5 shows a shock tube blast simulator with a wall specimen installed into it
There are few test standards for blast testing, but two commonly used standards are ASTM test standards for doors and windows. ASTM F1642, Glazing
and Glazing Systems Subject to Airblast Loadings, is described in Chapter 10.
It describes test setup, test execution, and performance standards for windows
subjected to blast loads or simulated blast loads. ASTM F2247, Standard Test
TESTING
Figure 5.4 Digital Photography and Deflection Measurement
Figure 5.5
Shock Tube Blast Simulator
155
156
PERFORMANCE VERIFICATION
Method for Metal Doors Used in Blast Resistant Applications, provides similar
standards for doors.
It is very important that test results undergo rigid verification, which includes
both validation against other test results and evaluating test reports. The latter is
particularly important. Somebody who understands test reports should carefully
review all of the conditions under which the tests were run, how the measurements were made, and how the test specimen responded. He or she must ensure
that the test was run in a valid manner and that the results are meaningful. In addition to evaluating the conduct and setup of the test, he or she must also verify
that the results are valid based on issues such as accuracy of the loading, whether
scaling was taken into effect correctly, and whether clearing was an issue or was
accounted for. This is less difficult for tests that are conducted in accordance with
a particular standard, but those tests must also be evaluated carefully. In many
ways explosive testing is more of an art than a science.
5.4 ANALYSIS
There are many ways to analyze the response of building elements to blast
shocks, ranging from approximations to highly detailed analyses such as finite
element analysis, and for each of those there are multiple computer codes for
performing that analysis. There are also hand methods and charts. Analysis is
the least expensive means for evaluating performance, but because of the nonlinearity of such analyses and the uncertainties in the loadings, dynamic material
properties, and dynamic structural response, analyses require stringent verification and validation.
Verification includes evaluating the applicability of specific methods to individual situations and ensuring that the models are applied correctly. Depending
on the assumptions upon which models are based, they may have limited applicability. Optimally, the applicability should be verified before they are applied, but
if not, their applicability must be verified in evaluating the results of the analyses.
While the applicability of a model to a given situation needs to be verified,
it is at least as important that models be validated. Models can be validated by
comparing their results to the results of other models that have been validated
for identical situations. For example, a computer code to be validated can be
used to analyze a beam, a wall, a window, or any other building component
of a specific geometry and configuration, and those results can be compared to
those derived by using a previously validated code. This is particularly the case
with finite element analyses, where there are so many variables that have to be
considered in the models that results can appear reasonable, but turn out to be
fallacious.
Results can also be compared to test results, which can serve as benchmarks.
In either case, validation gives the user both confidence in the model and confidence that its application is reasonable for that situation. This is particularly
useful for finite element models.
REFERENCES
157
In addition to analyzing buildings and building components to blast loads,
there are models for analyzing vehicle barriers. Because they model such
complex situations as vehicle–structure interaction, structural response, and
soil–structure interaction, those models tend to be very complicated. They often include finite element analysis models. Because they’re so complicated, it
is especially important that they be validated. Usually that validation is done
by either comparing them to existing test results or performing testing after the
analysis to validate the analysis results. The most common use of analysis models for vehicle barriers, therefore, is to design the barriers and validate the results
through testing.
5.5 PEER REVIEW
Because analysis and testing for blast loads involve such complexity, it is very
important that test results and analyses be validated through peer review. Peer
review is particularly important for structures with unique structural features or
structures that have high importance or high economic value.
In addition, because the number of experienced blast consultants is still limited, peer review is particularly important where analysis is done by less experienced designers.
Verification and validation through peer review can take multiple forms. As a
minimum, it should include detailed review of drawings to ensure that details are
consistent with the assumptions in the design analyses and that the drawingsare
properly detailed, to allow the expected performance to be achieved. Verification
should also evaluate the applicability of the models used and the accuracy with
which they were applied. It should also include checks of at least major elements
of the design calculations.
Peer review should be done by people who have significant experience in
applying the models used for the applications on which they were applied.
REFERENCES
ASTM. 2003. Standard Test Method for Metal Doors Used in Blast Resistant Applications (F2247). West Conshohocken, PA: ASTM International.
. 2004. Glazing and Glazing Systems Subject to Airblast Loadings (F1642).
West Conshohocken, PA: ASTM International.
. 2007. Vehicle Crash Testing of Perimeter Barriers (F2656). West Conshohocken, PA: ASTM International.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
II
Blast Phenomena and Loadings
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
6
Blast Phenomena
Paul F. Mlakar and Darrell Barker
6.1 INTRODUCTION
Blast is a pressure disturbance caused by the sudden release of energy. People
often think of blasts in terms of explosions such as the detonation of an explosive
charge. However, there are many other blast sources that have the potential to
cause damage. For example, chemicals may undergo a rapid decomposition under certain conditions. These events are often referred to as runaway reactions.
Flammable materials mixed with air can form vapor clouds that when ignited
can cause very large blasts. Blasts are not always caused by combustion; they
can also result from any rapid release of energy that creates a blast wave, such as
a bursting pressure vessel from which compressed air expands, or a rapid phase
transition of a liquid to a gas.
The loads resulting from a blast are created by the rapid expansion of the energetic material, creating a pressure disturbance or blast wave radiating away from
the explosion source, as shown in Figure 6.1. Blast pressure is more properly
overpressure, because it is relative to ambient conditions, rather than an absolute
pressure. Shock waves are high-pressure blast waves that travel through air (or
another medium) at a velocity faster than the speed of sound. Shock waves are
characterized by an instantaneous increase in pressure followed by a rapid decay. Pressure waves are lower amplitude and travel below the speed of sound.
Pressure waves are characterized by a more gradual increase in pressure than a
shock wave, with a decay of pressure much slower than a shock wave. In most
cases, shock waves have a greater potential for damage and injury than pressure
waves.
As a blast wave travels away from the source, the pressure amplitude decreases, and the duration of the blast load increases. Overexpansion at the center
of the blast creates a vacuum in the source region and a reversal of gas motion. This negative pressure region expands outward, causing a negative pressure
(below ambient), which trails the positive phase. The negative phase pressure
is generally lower in magnitude (absolute value) but longer in duration than the
positive phase. Generally speaking, positive phase blast loads are more consequential than negative phase loads, the latter of which is often ignored.
161
162
BLAST PHENOMENA
GROUND – REFLECTED WAVE
ASSUMED PLANE
WAVE FRONT
STRUCTURE
GROUND
SURFACE
W
Figure 6.1 Propagating Blast Wave
Expansion of the blast wave causes air particles to move outward during the
positive phase and inward during the negative phase. The flow of air particles
creates a pressure analogous to that caused by wind. The pressure produced by
this flow is referred to as the dynamic pressure. This pressure is lower in magnitude than the shock or pressure wave and imparts a drag load similar to wind
loads on objects in its path.
As the shock or pressure wave strikes a wall or other object, a reflection occurs, increasing the applied pressure on the surface. This reflected pressure is
considerably higher than the incident or free-field pressure wave. At the free
edges of a reflecting surface, the discontinuity between the forward traveling
incident blast wave and rearward traveling reflected blast wave creates a rarefaction, or pressure relief wave. Rarefactions travel inward from the outer edges
across the face of the reflecting surface. The rarefaction waves relieve the positive
reflected pressure down to the free-field or side-on pressure, plus drag pressure.
The peak reflected pressure is not affected, only the duration The time required
for the rarefaction waves to completely relieve the reflected pressure is termed
the clearing time. This clearing time varies across the surface. It should be noted
that a rarefaction wave does not instantaneously clear reflected pressure; rather,
the relief is somewhat gradual and takes longer than the time required for the
leading edge of the rarefaction to travel to the center of the reflected surface. If
the clearing time exceeds the positive phase blast wave duration, clearing does
not affect the positive phase loads.
6.2 SOURCES OF BLASTS
Blasts involving chemical reactions can be classified by their reaction rates
into two primary groups: deflagrations and detonations. A deflagration is an
SOURCES OF BLASTS
163
Pso
PRESSURE
POSITIVE IMPULSE, is
Ps(t)
t A+ t 0
NEGATIVE IMPULSE, is
tA
Po
Pso
tA
POSITIVE
PHASE
DURATION
NEGATIVE
PHASE
DURATION
t0
t 0–
Figure 6.2 Pressure Wave from Deflagration
oxidation reaction that propagates at a rate less than the speed of sound in the unreacted material. The corresponding blast wave is often termed a pressure wave
and has a finite rise time, as illustrated in Figure 6.2. A “fast” deflgration can
create a more sudden rise in pressure. By contrast, in a detonation, the reaction front propagates supersonically, usually many times faster than the speed of
sound. This blast wave is termed a shock wave and has an instantaneous rise in
pressure, as seen in Figure 6.3. Since pressure is closely related to reaction rate,
detonation pressures are usually many times higher than deflagration pressures.
A blast involving an explosive is an exothermic chemical reaction, usually involving an oxidizer and a fuel. Explosion reactions can be produced by a wide
array of materials, some familiar, such as trinitrotoluene (TNT), while others are
less well known. The reaction rate is dependent on the chemical and physical
properties of the energetic material, reactant proportions and homogeneity, geometry of the material, characteristics of the “container” in which the material
resides, method and energy of initiation, and other initial conditions. A deflagration may be initiated by a “soft” ignition source such as friction, spark, or open
flame. Under certain conditions, a deflagration can transition into a detonation.
Alternatively, detonations can be directly initiated in an explosive material that
exceeds certain minimum geometry constraints if it is impinged upon by a shock
source of sufficient strength.
Deflagrations and detonations may involve oxidizers and fuels that are
oxygen-deficient, or materials that may produce flammable gases as a product
of reaction. In either case, the unreacted or flammable products may mix with
air and result in secondary burning. The secondary burning does not contribute
164
BLAST PHENOMENA
Pso
PRESSURE
POSITIVE IMPULSE, is
NEGATIVE
IMPULSE, is–
Ps(t)
AMBIENT
Po
tA
tA+ to+ t –
o
tA+ to
–
Pso
tA
t=0
Ps–(t)
POSITIVE
PHASE
NEGATIVE PHASE
DURATION
to
DURATION
t–
o
TIME AFTER EXPLOSION
Figure 6.3 Shock Wave from Detonation
significantly to the blast pressures for external explosions but can be a major
consideration for predicting internal explosion blast pressures.
Industrial explosive, propellant, and pyrotechnic manufacturing provides a
wide array of energetic material configurations and materials. These products
may include gas generators for airbags, packaged general-purpose explosive
charges, shaped high-explosive charges for cutting, or blasting agents used in
mining. Some materials used in these manufacturing processes may not be well
characterized in terms of potential explosive potential or reactivity hazards.
Therefore the structural analyst should take care to understand the full range
of blast loads that may be produced, and to design and apply appropriate factors
of safety to account for uncertainties.
Terrorist threats principally involve solid materials such as plastic explosives
or improvised explosives such as ammonium nitrate and fuel oil (ANFO). These
materials can readily produce detonations, but the blast strength may be low
due to inefficient configuration or the presence of contaminants. Flammables
can potentially be used, but they are more difficult to transport and successively
initiate.
For blasts originating from a chemical reaction, the material involved and its
state, the proportions of fuel and oxidizer (if applicable), container strength and
configuration, and the method and energy of initiation are among the parameters
that determine blast characteristics. Explosive materials can be broadly categorized based on their state. Solid materials typically produce high-pressure, shortduration loads, while combustible gases of comparable energy produce lowerpressure, longer-duration loads. Energy output and standoff distance in the configuration of interest are key to accurately determining blast loads acting on a
structure. Blast sources are described by group in the following paragraphs.
SOURCES OF BLASTS
165
High explosives are materials that are intended to produce detonation events
with supersonic reaction fronts. The reactions proceeding through the explosive
are self-sustaining if the charge is of sufficient diameter and properly initiated.
Reaction rate, or detonation velocity, varies with material type and is a key factor in detonation pressure for a material. Detonation velocities typically range
from 3000 to 30,000 ft/s (1–9 km/s). High explosives can be classified by their
sensitivity to initiation into primary, secondary, and tertiary explosives. Primary
explosives are very sensitive to initiation—for example, lead azide used in detonators. Secondary high explosives include TNT, composition C4, and dynamite.
Tertiary explosives are much less sensitive to initiation and require a powerful
booster charge to initiate them. The blasting agent ANFO is an example of a
tertiary explosive.
For convenience in predicting blast pressures, the energy release of high explosives is commonly measured as a value relative to that of TNT. This “TNT
equivalence” is used to determine a TNT charge weight capable of producing the
same explosion energy, blast pressure, or blast impulse as the explosive of interest. The TNT equivalence is different for energy, peak pressure, and impulse, and
separate TNT equivalent energy values are reported for many materials. Equivalencies for a number of these materials appear in Table 6.1.
High-explosive detonations may be accidental or intentional events. Accidents
involving high explosives can occur in processing, handling, and transportation.
Intentional detonations can include explosives testing, military weapons, demolition, specialized cutting or explosive forming, and terrorist acts. In the case of
intentional detonations, structures, such as test structures, may be required to
withstand multiple events.
Confined explosive charges will create gas pressures in the structure in addition to shock waves. If the structure is not adequately vented, the effects of these
gas pressure loads may exceed those of the shock loads.
High explosives come in many forms determined by their chemical and physical properties and intended use. Most explosives have additives to aid in stabilizing the material chemically or physically. Powdered explosives may be combined
with a plastic binder to form plastic-bonded or PBX explosive. Extrudable explosives have a viscous form due to the addition of rubber resins to the explosive
material. This addition allows the explosive to be molded or extruded to produce
a particular shape.
Certain liquids, such as nitromethane, can also detonate. Nitromethane is relatively insensitive and must be initiated by a strong ignition source such as a
high-explosive booster. It does have safety advantages over other materials and
may be handled as a simple flammable liquid in many cases. The shock front
produced by nitromethane is well formed and produces blast loads with approximately 100% TNT equivalence.
Propellants and pyrotechnics, also known as low explosives, are energetic materials that do not typically detonate and are used to produce gas, smoke, flash,
or sound. Both solid and liquid rocket propellants, gun propellants, and black
powder are examples of low explosives. Pyrotechnics include many different
166
BLAST PHENOMENA
Table 6.1
TNT Equivalence of High Explosives
Explosive
Density
Mg/m3
Equivalent
Moss for
Pressure
Equivalent
Mass for
Impulse
Pressure
Range MPa
Amatol (50/50)
1.59
0.97
0.87
NA1
Ammonia Dynamite
(50 percent Strength)
1
0.90
2
NA
0.90
NA1
Ammonia Dynamite
(20 percent Strength)
NA1
0.70
0.702
NA1
ANFO (94/6 Ammonium
Nitrate/Fuel Oil)
NA1
0.87
0.872
0.03 to 6.90
AFX-644
1.75
0.732
0.732
NA1
1.59
2
2
NA1
2
AFX-920
1.01
2
1.01
AFX-931
1.61
1.04
1.04
NA1
Composition A-3
1.65
1.09
1.07
0.03 to 0.35
Composition B
1.65
1.11
1.20
0.98
1.30
0.03 to 0.35
0.69 to 6.90
Composition C-3
1.60
1.05
1.09
NA1
Composition C-4
1.59
1.20
1.37
1.19
1.19
0.07 to 1.38
1.38 to 20.70
Cycloid (75/25 RDX/TNT)
(70/30)
(60/40)
1.71
1.73
1.74
1.11
1.14
1.04
1.26
1.09
1.16
NA1
0.03 to 0.35
NA1
DATE
1.80
0.87
0.96
NA1
2
Explosive D
1.72
0.85
0.81
0.01 to 0.30
Gelatin Dynamite
(50 percent Strength)
NA1
0.80
0.802
NA1
Gelatin Dynamite
(20 percent Strength)
NA1
0.70
0.702
NA1
H-6
1.76
1.38
1.15
0.03 to 0.70
HBX-1
1.76
1.17
1.16
0.03 to 0.14
HBX-3
1.85
1.14
0.97
0.03 to 0.17
HMX
NA1
1.25
1.252
NA1
1.80
2
NA1
L.X-14
1
NA
1.80
MINOL II
1.82
1.20
1.11
0.02 to 0.14
Nitrocellulose
1. 65 to 1. 70
0.50
0.502
NA1
Nitroglycerine Dynamic
(50 percent Strength)
NA1
0.90
0.902
NA1
Nitroglycerine (NQ)
1.72
1.00
1.002
NA1
Nitromethane
NA1
1.00
1.002
NA1
(continued)
SOURCES OF BLASTS
167
Table 6.1 (Continued)
Explosive
Density
Mg/m3
Equivalent
Moss for
Pressure
Equivalent
Mass for
Impulse
Pressure
Range MPa
Octol (75/25 HMX/TNT)
(70/30)
1.81
1.14
1.02
1.09
1.06
1.092
NA1
0.01 to 0.30
PBX-9010
1.80
1.29
1.292
0.03 to 0.21
PBX-9404
1.81
1.13
1.70
1.132
1.70
0.03 to 0.69
0.69 to 6.90
PBX-9502
1.89
1.00
1.00
NA1
2
PBXC-129
1.71
1.10
PBXN-4
1.71
0.83
PBXN-107
1.64
PBXN-109
1.67
1
1.05
2
1.05
2
1.10
NA1
0.85
NA1
2
NA1
2
NA1
2
1.05
1.05
PBXW-9
NA
1.30
1.30
NA1
PBXW-125
1.80
1.022
1.022
NA1
Pentolite. (Cast)
1.64
1.68
NA1
1.42
1.38
1.50
1.00
1.14
1.00
0.03 to 0.69
0.03 to 4.14
0.69 to 6.90
PETN
1.77
1.27
1.272
0.03 to 0.69
Picroto1 (52/48 Ex D/TNT)
1.63
0.90
0.93
0.03 to 4.10
RDX
NA1
1.10
1.102
NA1
RDX/Wax (98/2)
1.92
1.16
1.162
NA1
1.30
2
NA1
2
NA
1
TATB
NA
1
1.00
1.00
NA1
Tetryl
1.73
1.07
1.072
0. 02 to 0.14
2
RDX/AL/Wax (74/21/5)
1.30
1.59
1.06
TNETB
1.69
1.13
0.96
0.03 to 0.69
1.75
1.18
123
1.23
1.32
1.38
1.11
1.322
1.382
0.03 to 0.69
NA1
NA1
1.63
1.00
1.00
Standard
TNETB/AL
(90/10)
(78/22)
(65/35)
TNT
1.06
NA1
Tetrytol (75/25 Tetryl/TNT)
Torpex
1.85
1.23
1.28
0.01 to 0.30
Tritonal (80/20 TNT/AL)
1.72
1.07
0.96
0.03 to 0.69
1
2
NA - Data not available.
Value is estimated.
materials such as fireworks, smoke grenades, flares, and the gas generator material in automobile air bags. The relatively low reaction rates of these materials
result in low blast pressure output because they typically don’t detonate. Fire
hazards and gas generation concerns dominate the safety analysis in most cases,
but detonations or deflagrations cannot be ruled out in all cases.
168
BLAST PHENOMENA
Figure 6.4 Vapor Cloud Explosion in 9/11 Pentagon Crash
Predicting blast loads for these materials can be challenging. There is much
less data available for propellants and pyrotechnics than for high explosives, with
respect to blast loads. Burn rates, gas generation, and energy output rates are
available in handbooks for common propellants (Cooper and Kurowsky 1996)
These values can be used to perform a burn simulation and determine the gas
pressure load in a confined area. In some cases, TNT equivalencies for these
materials have been used. However, this is problematic because peak pressure
for propellants and pyrotechnics is much lower than for TNT, but the impulse
of propellents can be higher in a confined space than for the TNT equivalent.
Care must be taken when equivalencies are based on small-scale tests, as large
quantities may react in a more energetic manner.
Vapor cloud explosions involve the release of a flammable material which,
when mixed with air and given the proper conditions, forms an ignitable material, potentially leading to an explosion. The vapor cloud explosion following the
9/11 Pentagon crash is shown in Figure 6.4. Only fuel within flammability limits
participates in the explosion, which is a minor percentage of the total quantity of
fuel released (Baker et al. 1991). Much of the fuel vapor is too lean or too rich to
combust and does not produce blast overpressure. Blast output of a vapor cloud
explosion is highly variable and depends on the fuel, size of the flammable vapor cloud, presence of congestion (obstacles such as structural members, pipes,
and vessels) in the flammable region, and presence of confining surfaces. Release of a flammable material into the open air, absent congestion or confinement, typically results in a fireball that does not produce damaging overpressures.
SOURCES OF BLASTS
169
Confinement and congestion in the flammable cloud create turbulence and accelerate the combustion, thereby increasing the flame speed. Peak overpressure is
a strong function of flame speed; thus, a key element in vapor cloud explosion
prediction is maximum flame speed. Vapor cloud detonations are rare and only
known to occur with high-reactivity materials, strong ignition sources, and/or
severe cases of confinement and congestion. The vast majority of vapor cloud
explosions are deflagrations. A particularly noteworthy characteristic of deflagrations is that high flame speeds only occur in confined/congested regions of
the flammable cloud. Outside of congested areas, flame speed decreases. As a
result, deflagration blast loads are produced in confined/congested areas, rather
than at the point of release or dispersion. In the rare case of vapor cloud detonations, the blast loads are produced by the detonatable material and are typically
considered to be centered in the dispersed cloud and not at the point of release.
Overpressures produced by vapor cloud explosions are substantially lower
than those produced by high explosives. For example, a vapor cloud detonation may reach 16–18 atmospheres overpressure, compared with 100,000 atmospheres or greater for most high explosives. However, the energy of vapor cloud
explosions can be very large, as evidenced by some of the devastating accidents
which have occurred. In Flixborough, U.K., a 30,000-kg release of cyclohexane
caused an explosion that resulted in 28 fatalities, multiple injuries, and destruction of more than 150 structures. The large energy in hydrocarbons produces a
commensurately large impulse. Blast predictions made using a TNT equivalent
approach tend to be inaccurate due to the disparity between the pressure and
impulse compared with high explosives. As a result, TNT equivalence analyses
of vapor cloud explosions are not recommended. A more accurate method is to
perform a dispersion analysis of the flammable release to determine fuel in the
flammability limits, select appropriate speed or reaction severity, then use nondimensional scaled energy curves to determine the pressure and impulse at various standoff distances. Computational fluid dynamics (CFD) analysis is another
method of determining blast pressures for vapor releases.
Vessels designed to contain fluids under pressure have the potential to create
hazardous overpressures and fragments if the vessel fails while under pressure. A
pressure vessel burst may occur at normal working pressure due to a mechanical
integrity problem, weakening, or other external factors (such as impact); or at
elevated pressure due to excessive pressurization from an external source, an
uncontrolled internal chemical reaction, heating vessel contents with insufficient
pressure relief, or other means. This type of accident results in loss of contents
and may lead to a follow-on explosion when the contents subsequently mix with
air, if the contents are flammable or pyrophoric.
In the far field, the blast load from a vessel that fails at high pressures (several
thousand psi [1000 psi = 6895 kPa] or higher) is similar to that caused by high
explosives. In the near field, overpressures are less than those produced by high
explosives, and the use of TNT equivalent prediction procedures will overestimate pressure. Key parameters for estimating blast effects from bursting vessels
170
BLAST PHENOMENA
are the volume of the material in the vessel, the failure pressure, the composition
of the contents, and the temperature of the contents.
Fragments represent an additional hazard from bursting vessels. Failure at
working pressure or overpressurization at a relatively slow rate usually results
in a small number of fairly large fragments. In some cases, one or both ends
of a cylindrical pressure vessel may be propelled long distances by a rocketing
mechanism. Large fragments may be thrown as a result of the expansion of the
vessel contents. Containment of these fragments, which can be quite challenging,
may be required to provide adequate safety. Fragment hazard prediction methods
are discussed further in Chapter 8 of this handbook.
A boiling liquid expanding vapor explosion (BLEVE) occurs when a vessel
containing a liquid stored above its boiling point for atmospheric pressure fails
and a rapid depressurization of the liquid occurs. Upon depressurization, a portion of the superheated liquid rapidly boils, and the expanding vapors contribute
to the blast wave. Compressed gases in the vessel prior to failure also contribute
to the blast wave as in a bursting pressure vessel. A common BLEVE scenario
is a pressure vessel engulfed in an extended fire, causing weakening of the vessel shell to the point of failure. A BLEVE may involve a flammable liquid, and
mixing of the flammable material with air and subsequent ignition may cause a
fireball. A detailed discussion of BLEVE events and of simplified blast prediction methods is provided in CCPS’s Guidelines for Evaluating the Characteristics of Vapor Cloud Explosions, Flash Fires, and BLEVES (Center for Chemical
Process Safety 1994).
6.3 CHARACTERISTICS OF BLAST WAVES
6.3.1 Key Parameters
An idealized pressure-time history indicating the key parameters of a blast load is
shown in Figure 6.3. Figure 6.5 shows a comparison between free-field, or sideon, and reflected pressure-time histories. The parameters shown in these figures
are defined as:
Po
Pso
= Ambient pressure
= Peak positive side-on overpressure
Pso−
Ps (t)
= Peak negative side-on overpressure
= Time varying positive overpressure
Ps− (t) = Time varying negative overpressure
Pr
= Peak reflected overpressure
Is
= Positive-phase-specific impulse, the integration of the positive
phase pressure-time history
−
= Negative-phase-specific impulse, the integration of the negative
is
phase pressure-time history
PRESSURE
CHARACTERISTICS OF BLAST WAVES
171
Pr
Pso
to+ t –
o
to
Po
POSITIVE
NEGATIVE PHASE
PHASE
DURATION, to
DURATION, t o–
TIME
Figure 6.5 Comparison of Free Field and Reflected Blast Loads
ta
to
= Time of arrival
= Positive phase duration
to− = Negative phase duration
The term “overpressure” refers to a gauge pressure, or the blast pressure relative to ambient pressure.
Free-field loads are those produced by blast waves sweeping over surfaces
unimpeded by any objects in their path. This load is also referred to as side-on
when the blast wave sweeps over a wall or other object parallel to its direction
of travel. Side-on pressure terms are indicated by a “so” subscript in the figures
in this chapter, as well as in most references.
6.3.2 Scaling
Blast pressures, load duration, impulse, shock wave velocity, arrival times, and
other blast parameters are frequently presented in scaled form. Research has
172
BLAST PHENOMENA
shown that scaling laws can be applied to explosions, allowing data from one
explosion test to be applied to a geometrically similar (larger or smaller) case.
As a result, scaling has tremendous utility in blast predictions, allowing compilations of explosion test and numerical modeling data to be used to predict loads
for any combination of explosion energy and standoff distance within the range
of data. The most common form of scaling is called “cube root scaling” owing to
the fact that blast parameters are scaled by the cube root of the explosion energy.
Both dimensional and dimensionless scaling are used, and care must be taken
with all unit parameters to ensure the scaling is correctly applied for the blast
curves being used.
6.4 PREDICTION OF BLAST PARAMETERS
This section describes the principal methods for predicting blast parameters, with
some discussion of practical applications and limitations. A more detailed treatment is given for high explosives and vapor cloud explosions because these tend
to be the most common design threats for protective construction. Definition of
the blast hazard is discussed in Chapter 2, and calculation of the loads imposed
on the structure in Chapter 7.
In many cases, blast load prediction is a multi-step process. Often an initial
estimate of blast loads will be made using simplified techniques and a conservative estimate of explosion energy. It may become obvious at this point that
the predicted loads are so low they are not a contributing factor in structural design, or will not cause undue cost or complexity in the design. Thus, the initial
loads are judged to be adequate for analysis or design. If loads are high enough
to drive the design, it may be appropriate to perform a more detailed blast load
prediction which uses less conservative estimates of energy and more detailed
prediction techniques.
The basic simplified approach involves predicting free-field loads using empirical or semi-empirical methods. Blast curves that relate explosion energy,
standoff distance, and blast wave parameters are typically used in this initial
load-prediction step. These curves are readily available for high explosives and
vapor cloud explosions.
6.4.1 High Explosives
A significant amount of data has been produced that quantifies the relationships
among charge weight, standoff distance, and blast parameters for both positive
and negative phases. A Unified Facilities Criteria publication, Structures to Resist the Effects of Accidental Explosions, produced by the Department of Defense, contains the relationships in the form of scaled blast curves (U.S. Department of Defense 2008). This document is publicly available, and an electronic
version also exists to aid in the use of the numerous tables and figures.
PREDICTION OF BLAST PARAMETERS
173
The empirical blast parameter curves provided in this document plot air-blast
parameters versus scaled distance for both air-burst and surface-burst configurations. A charge detonated on the ground will produce higher blast loads than an
air burst due to the reflection of the shock by the ground surface. For unyielding surfaces with the charge located on the surface, the effect is a doubling of
the charge weight, since the energy of the blast directed to the ground is fully
reflected. For soft soils or charges located above the surface, the reflection factor
is less than two, since some of the energy is absorbed by the substrate. Some
guidance for selecting the reflection factor is given in the document, but it is
conservative to assume a fully reflecting surface.
The most commonly used approach to present blast wave relationships for
high explosives is the Hopkinson-Cranz, or cube root, scaling method. Figure 6.6
is a simplified version of TNT blast curves that provides the parameters for
charges located at ground level. The cube root term results from geometric scaling laws in which charge diameter varies in proportion to all distances, and thus
the charge weight is proportional to the cube of the charge diameter. To use these
empirical curves, one computes the scaled distance by dividing the standoff distance from the charge to the point of interest by the cube root of the charge
weight. For explosives, this takes the form of
Z
Z
R
W
= R/W 1/3 where :
= scaled distance (ft/lb1/3 )
= standoff distance (ft)
= explosives weight (lb)
(6.1)
The curves provide pressures, P, which are the same at a given scaled distance. The curves also provide scaled times, t/W 1/3 , and scaled impulses, I/W 1/3 .
The actual times and impulses are then found by multiplying the scaled values
by W 1/3 .
In accordance with the common practice for blast design, the variables in
Equation 6.1 and Figure 6.6 have physical dimensions. However the procedure
is rigorously founded on dimensionless modeling. It is also based on atmospheric conditions at sea level. While this suffices for most practical situations,
the effects of differing ambient conditions can be treated more comprehensively
through Sachs or energy scaling (Baker 1973).
As an example of cube root scaling, let us consider the Oklahoma City Bombing. This event has been reported to have been equivalent to the detonation of
4,000 lbs of TNT at essentially the ground surface. If a location of interest is
100 ft away, the scaled distance is
Z = 100/(4000)1/3 = 6.30 ft/lb1/3
From Figure 6.6 we have an incident pressure of Po = 24.9 psi and a reflected
pressure of Pr = 79.5 psi at this scaled range. This figure also provides the scaled
positive phase duration to /W 1/3 = 1.77 msec/lb1/3 . From this, the total positive
174
BLAST PHENOMENA
Figure 6.6 Blast Parameters for TNT Surface Bursts
phase duration is
to = 1.77 × (4000)1/3 = 28.1 msec
The figure similarly indicates the scaled positive-phase-specific impulse to be
Is = 12.17 psi × msec/lb1/3 and the total impulse is
Is = 12.17 × (4000)1/3 = 193.2 psi × msec
Some structures, such as explosives or propellant processing bays, mail
rooms, and loading docks, often require blast design for containment of internal
PREDICTION OF BLAST PARAMETERS
175
explosion effects. Containment designs must consider that shock waves emanate
from the charge, strike wall and roof components, and reflect to impact other
surfaces. Each surface is subjected to multiple shock waves.
Internal explosions also produce gas and heat, which cause a pressure increase
within the containment. The peak gas pressure, Pg , developed is a function of
the charge weight/free volume ratio in the containment. This pressure buildup is
relatively slow compared to the load duration associated with shock waves. For
simplicity, the peak gas or quasi-static pressure is assumed to rise instantaneously
to the peak gas pressure but is not additive to shock pressures. A curve of peak
gas pressuresversus weight/volume is shown in Figure 6.7.
For example, let us consider the detonation of W = 200 lbs of TNT within
a structure having a free volume of Vf = 50,000 ft3 . This loading density is
W/Vf = 200/50,000 = 4 × 10−3 lb/ft3 . From Figure 6.7 the peak gas pressure is
Pg = 47.2 psi.
The duration of the gas pressures is a function of the vent area available and
the rate at which temperature decays in the containment area. Vent area can
change over time, depending on how fast the vent cover moves away from the
enclosure. The vent velocity is a function of the applied shock impulse and the
frangibility and mass of the vent cover. The duration is determined by first computing the peak gas pressure from Figure 6.7. Then the gas impulse, Ig , is estimated from a set of graphs as a function of the scaled vent area, scaled mass of
the vent cover, loading density, and scaled shock impulse (U.S. Department of
Defense 2008). An equivalent triangular gas duration is then
tg = 2Ig /Pg
Analytical methods have been developed to predict blast loads. These methods fall into two groups: semi-empirical and hydrocode. The semi-empirical approach utilizes a physics-based model to compute selected blast parameters with
coefficients that are “tuned” to match test data. These models are limited to configurations and charge weight ratios for which data are available, but offer the
advantage of quick run times compared with more detailed techniques. Semiempirical codes may offer limited abilities to model shock diffraction, shielding, and reflection. Semi-empirical methods have been developed primarily by
defense-related agencies and are restricted to distribution to the government and
its contractors.
Hydrodynamic codes, or hydrocodes, utilize a grid of computational cells to
track detonation propagation through an explosive charge and shock wave propagation through a medium. Using principles of material behaviors, fundamental thermodynamics, and fluid dynamics, hydrocodes predict pressure, density,
and other key parameters. This type of analysis is much more complex than
the use of empirical relationships. Some of the available tools have user interfaces that greatly simplify the analysis, especially for standard materials, but
considerable effort and expertise are typically required to conduct a competent
analysis.
176
BLAST PHENOMENA
Figure 6.7 Peak Gas Pressure for Internal Detonations
The output from hydrocodes will be in the form of pressure-time series for
selected locations in the model. Application of these loads to the structural
model can be tedious. If the structural analysis involves a time series calculation,
the time varying loads may be transferred into the structural analysis, but the
number of data points to be transferred may be excessive. Normally, a simplification is made to the applied load history to greatly reduce the number of load
points to a series of lines approximating the “real” curve. This reduced multipoint curve is more complex than the linear decay curve used for the empirical
relationship analysis. A coupled analysis may be performed in the hydrocode
PREDICTION OF BLAST PARAMETERS
177
Figure 6.7 (Continued)
which models blast propagation and structural response. This approach offers
additional capability for detailed response analysis but with significant increase
in complexity compared with simulation of blast loads alone.
6.4.2 Bursting Pressure Vessels
Bursting pressure vessel blast predictions are most commonly conducted in one
of two ways: simplified analytical methods or computation fluid dynamics (CFD)
178
BLAST PHENOMENA
models. In either case, it is necessary to select the conditions under which the
vessel would be expected to fail—overpressure/overfilling, runaway reaction, external heating, mechanical or impact load, corrosion, or other mechanisms. This
selection is important because it determines the composition, pressure, temperature, phase, and energy of the vessel contents, all of which can have a significant
effect on the predicted blast loads. In any case, as long as the vessel pressure
is above two atmospheres at failure and the pressure is released abruptly, the
resultant blast will be a shock wave (i.e., suddenly applied).
The most recent simplified method was published by Tang (Tang et al. 1996)
and consists of a series of nondimensionalized curves to predict the side-on pressure and impulse at a particular standoff as a function of pressure ratio (defined
as burst pressure divided by ambient pressure), ratio of specific heats of the vessel contents (a property of the gaseous contents of the vessel), expansion energy
potential of the vessel contents, and the temperature ratio (the ratio of the vessel
internal temperature at failure to the ambient temperature). The vessel energy
term in this analysis should be corrected for ground reflection, liquid fraction in
the vessel, and energy consumed by fragment production. In addition, the sideon pressures need to be modified to account for reflection and clearing effects,
as described in the next chapter.
If more detail is necessary, a CFD model can be applied. These consider the
thermodynamics and gas properties to compute the production and propagation
of blast waves produced by a bursting vessel. It is important to note that there
are complex interactions between the failing vessel and the fluid escaping the
vessel that cannot be adequately modeled by any publicly available current program. In addition, the failure of the vessel is not wholly deterministic, as cracks
can propagate along micro-cracks and grain boundaries in a manner that cannot
usually be predicted. As a result, these numerical analyses will rely heavily on
the experience and expertise of the analyst. The advantage of using a hydrocode
is that it can directly produce load-time distributions on arbitrary surfaces and
account for shielding and focusing effects produced by adjacent structures.
6.4.3 Vapor Cloud Explosions
Prediction of blast loads for vapor cloud explosions can be more complex than
loads for high explosive detonations or bursting pressure vessels. The first step
in determining blast load from vapor cloud explosions is to develop the release
scenario for the flammable material. This selection process is typically done with
input from process engineers and plant operations. Dispersion modeling is performed to model the release scenario and determine the size, shape, and concentration of the flammable vapor cloud. This dispersion is overlaid onto areas of
confinement and congestion to define the flammable mass that may participate
in the explosion. In some cases, cloud dispersion is not evaluated, and the confined/congested volume is conservatively assumed to be filled with a stoichiometric mix of flammable gas. This approach may be justified if it is evident that
credible release scenarios can fill the volume. Blast loads may be determined
PREDICTION OF BLAST PARAMETERS
179
using simplified methods consisting of graphs (blast curves) of pressure and
impulse, or of duration versus scaled standoff, or through more complex modeling using CFD.
In the blast curve method, the scaled standoff is computed by using distance
from the center of the explosion to the point of interest and the energy content of
the confined/congested flammable mass. Scaled pressure and impulse values are
read from blast charts containing flame speed curves. These curves range from
low flame speeds that produce almost negligible blast pressure to the supersonic
flame speed of a detonation producing the highest pressure. Determination of
maximum flame speed requires knowledge of the amount of confinement and
congestion in the flammable portion of the vapor cloud, and the fuel reactivity.
The two most commonly used methods are the Baker-Strehlow-Tang (BST)
(Tang and Baker 1999) and TNO Multi-energy Method (MEM) (Van den Berg
1985). Figure 6.8 shows the overpressure, and Figure 6.9 shows the impulse
as a function of scaled distance and flame speed for BST. To use the curves,
compute the scaled distance Rbar = R(po /E)1/3 in which R = standoff distance,
po = ambient pressure, and E = heat of detonation. Then obtain the scaled pressure Pbar = (P/po ) from Figure 6.8 and unscale to determine the pressure P.
Next, determine the scaled impulse ibar = i ao /(E po )1/3 in which i = impulse and
ao = ambient sound velocity. The impulse, i, follows by unscaling. The equivalent triangular duration is then to = 2i/P.
The process is similar for MEM, using Figures 6.10 and 6.11 for pressure and
duration respectively. Impulse is i = Pt0 /2.
50
Mf=5.2
Mf=4.0
Mf=3.0
Mf=2.0
Mf=1.4
Mf=1.0
Mf=0.7
Mf=0.35
Mf=0.2
10
(P-P0)/P0
1
0.1
0.01
0.002
0.1
1
1/3
(R/E/P0)
Figure 6.8 Peak Pressure for Vapor Cloud Explosions—BST
10
180
BLAST PHENOMENA
Mf=5.2
Mf=4.0
Mf=3.0
Mf=2.0
Mf=1.4
Mf=1.0
Mf=0.7
Mf=0.35
Mf=0.2
i+a0/(E1/3 p02/3)
0.1
0.01
0.001
0.1
1
1/3
(R/E/P0)
10
Figure 6.9 Impulse for Vapor Cloud Explosions—BST
1E+02
10
9
8
7
6
5
4
3
2
1
1E+01
Scaled Pressure (P-P0)/P0
1E+00
1E-01
1E-02
1E-03
1E-04
1E-05
1E-06
1E-07
0.1
1
10
100
1000
Scaled Distance (R(P0/E)1/3)
Figure 6.10 Peak Pressure for Vapor Cloud Explosions—TNO MEM
10000
REFERENCES
181
10
Scaled Duration (t*a*(P0 /E)1/3)
1
2
3
4
1
5
6
7
8
9
10
0.1
0.1
1
10
100
1000
10000
Scaled Distance (R(P0/E)1/3)
Figure 6.11 Duration for Vapor Cloud Explosions—TNO MEM
A different approach is to develop a CFD model of the process area. A grid
of computational cells is constructed, and the controlling combustion chemistry
and gas kinetics equations are solved to track the progression of the reaction and
resulting pressures. The end result is a pressure-time series at selected points.
It is important for the structural analyst to know if the pressure-time plots are
free-field or applied loads. An advantage of this approach is that it allows the
analyst to include the effects of blockage and impingement of objects, which can
substantially influence the dispersion and explosion effects. This capability is not
available in the simplistic methods using scaled energy curves.
6.5 SUMMARY
Blast phenomena are characterized by a rapid release of energy resulting in a
pressure wave that propagates outward from the source. The source can be a high
explosive, propellant, or other gases that are released suddenly. Shock waves
are characterized by an instantaneous rise in pressure from ambient. Pressure
waves, are produced by deflagrations resulting from combustion of low explosives and flammable gases. The characteristics of a blast waves can be predicted
with relatively simple empirical methods or more complex computational fluid
dynamics. Once free-field or side-on blast pressure and impulse are defined, the
applied loads on structures can be determined, as described in the next chapter.
182
BLAST PHENOMENA
REFERENCES
Baker W. E. 1973. Explosions in Air. Austin: University of Texas Press.
Baker, Q.A. and W. E. Baker. 1991. Pros and cons of TNT equivalence for industrial
explosion accidents. Presented at AIChE International Conference on Workshop on
Modeling and Mitigating the Consequences of Accidental Releases of Hazardous Materials, New Orleans, LA, May 21–24, 1991.
Center for Chemical Process Safety. 1994. Guidelines for Evaluating the Characteristics
of Vapor Cloud Explosions, Flash Fires, and BLEVES. New York: American Institute
of Chemical Engineers.
Cooper, W. P. and R. S. Kurowski. 1996. Introduction to the Technology of Explosives.
New York: Wiley-VCH.
Tang M. J. and Q. A. Baker. 1999. A new set of blast curves from vapor cloud explosion.
Process Safety Progress 18 (4): 235–240.
Tang, M. J., C. Y. Cao, and Q. A. Baker. 1996. Calculation of blast effects from bursting vessels.” Presented at the 30th Annual AIChE Loss Prevention Symposium, New
Orleans, LA, February 1996.
U.S. Department of Defense. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense.
Van Den Berg, A. C. 1985. The multi-energy method—A framework for vapor cloud
blast prediction.” Journal of Hazardous Materials, 12 (1): 1–10.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
7
Blast Loading
Paul F. Mlakar and William Bounds
7.1 INTRODUCTION
In the previous chapter, blast phenomena and the characteristics of blast waves
were discussed. In this chapter, methods for determining the dynamic blast loading on structural elements from these characteristics are discussed. Emphasis is
placed on simple approximate procedures that apply widely in practical blastresistant design. More comprehensive methods are also discussed.
7.2 EMPIRICAL METHOD
Although research into the effects of explosions dates back to 1870, most development to determine the blast loading on buildings and other similar structures was started in the 1950s and 1960s by the U.S. military. Several publicly
available military manuals were distributed during this period (TM 5-856 and
TM 5-1300) that presented empirically-derived charts and equations. Several papers and publications published during that period (Newmark 1956, Biggs 1964,
ASCE 1985) also provided information for design.
The empirical method consists of published equations, graphs, tables, and
figures that allow one to determine the principal loading of a blast wave on a
building or a similar structure. Software has also been developed to automate
calculations based on this same source information. The basic advantage of empirical methods is speed and simplicity. Applicable factors and coefficients can
be looked up in tables, blast loading can be calculated manually, and a building
blast load can be determined in a matter of minutes. Empirical methods are also
well suited for situations where the accuracy of the blast source location and
magnitude is uncertain. More comprehensive methods, such as computational
fluid dynamics (CFD), require specialized software, operator training, and, potentially, weeks of data input and verification. The disadvantage of empirical
methods is the lack of flexibility in application. Most data are based on plain
rectangular target structures located in open terrain. Explosions are assumed to
be either an air blast or surface blast. The blast wave is typically based on high
183
184
BLAST LOADING
explosives producing a shock wave, as opposed to a deflagration that produces a
pressure wave. Empirical methods remain popular for the preliminary evaluation
of complex design situations and for the design situations where the location and
magnitude of the blast source are not well understood. The analyst must make a
judgment to pursue an efficient design calculation, and to balance the accuracy
of input information and the desired accuracy of results.
The most extensive and widely referenced publication for empirical design is
UFC 3-340-02 (formerly TM5-1300) (U.S. Department of Defense 2008). This
design manual addresses accidental explosions related to munitions manufacturing, handling, and storage. Nevertheless, many of the procedures are applicable
to buildings and structures designed for other blast scenarios. Basic information from UFC 3-340-02 is provided in this document for external loading. UFC
3-340-02 also provides load determination information on other configurations,
such as partially open cubicles and interior explosions. Information on empirical
methods is available from a number of other sources such as Mays and Smith’s
Blast Effects on Buildings (Mays and Smith 1995), Biggs’s Introduction to Structural Dynamics (Biggs 1964), and ASCE’s Structural Design for Physical Security and Design of Blast Resistant Buildings in Petrochemical Facilities (ASCE
Physical Security 1999, ASCE Petrochemical 1997).
The empirical blast load example calculations provided in this chapter assume
a blast wave interaction with a rectangular structure of finite size, such that the
structure is blast-loaded on all sides. Furthermore, due to the structure depth, or
building dimension parallel to the direction of the blast wave, there will be a significant net lateral force applied to the structure. Mays and Smith describe this
as one of three blast loading situations to keep in mind. The second situation is
a blast wave interacting with a relatively small structure, such as a vehicle, that
is effectively engulfed with blast pressure acting on all sides of the structure at
once. The third situation is for a blast wave acting on a relatively large structure,
such as a large office building, where the magnitude of the blast wave varies significantly across the surfaces of the structure. Some surfaces of these structures
may see little if any external blast loading.
Empirical methods typically rely on a straight-line equivalent of the actual
blast pressure-time relationship, as well as the use of a straight-line equivalent
loading on the structure of interest. The approximation is fairly close, and the
resulting calculations are much simpler.
Details and examples illustrating implementation of empirical methods are included in following sections of this chapter. Key input parameters for the determination of building loads, as presented in Chapter 6, are: the peak side-on overpressure, Pso , the positive phase duration, to , and the shock front velocity, Us .
One other parameter necessary for the determination of building blast loads
is the dynamic wind pressure, qo . Dynamic wind is the movement of air particles resulting from a shock wave. The effect is additive to the blast overpressure
and is a function of the free-field blast overpressure and the obstruction’s shape.
Open-frame structures and small buildings where a blast wave will produce quick
envelopment are most sensitive to dynamic wind. Values of qo can be determined
EMPIRICAL METHOD
185
Figure 7.1 Peak Incident Pressure versus Peak Dynamic Pressure, Density of Air
Behind the Shock Front, and Particle Velocity (UFC 3-340-02)
from Figure 7.1. Alternatively, in the low overpressure range, and at sea level
atmospheric pressure, the following equation from Newmark can be used.
qo = 0.022 (Pso )2
(7.1)
The pressure exerted on a structural element is the dynamic wind pressure
multiplied by a drag coefficient. The drag coefficient, Cd , is a function of the
shape and orientation of the obstructing element. Newmark lists approximate
values of Cd for open-frame structural elements as 2 for structural shapes, 1.25
for box shapes, and 0.8 for cylinders. Values of Cd for enclosed rectangular buildings are provided in the following sections.
186
BLAST LOADING
7.2.1 Empirical Method—Basic Blast Wave Example
The example calculation presented in the following sections is based on the blast
wave and building parameters described below. The given blast wave will be applied normal to the long side of a rectangular flat-roof building. It is further determined that the distance to the explosion and the length of the building are such
that the overpressure and duration do not change significantly over the length of
the building.
Given charge weight, W = 5 tons (TNT) = 10,000 lb:
The charge weight, W , should include any applicable safety factor
Given distance from source to building, R = 215 ft
The blast wave parameters, computed using Figure 6.6 as follows:
To use Figure 6.6, compute the scaled distance:
Z = R/W 1/3 = (215 ft)/(10,000 lb)1/3 = 10.0
Using Z = 10, use Figure 6.6 to determine the following parameters:
Peak side-on overpressure, Pso = 10 psi
Shock front velocity, Us = 1.4 ft/ms
Scaled positive phase duration, to /W 1/3 = 2.6 ms/lb1/3
Compute the positive phase duration:
to = (to /W 1/3 )W 1/3 = (2.6 ms/lb1/3 )(10,000 psi)1/3 = 56 ms
Compute the positive phase wave length:
L w = (Us )to = (1.4 ft/ms)(56 ms) = 78 ft
With Pso = 10 psi, use Figure 7.1 to determine the peak dynamic wind pressure, qo = 2.1 psi
Alternatively, use Equation 7.1 to calculate the peak dynamic wind pressure,
qo = 0.022 (Pso )2 = 0.022 (10 psi)2 = 2.2 psi
Loads are to be computed on an aboveground rectangular building with the
following dimensions,
Building height, B H = 20 ft
Building width, BW = 150 ft
Building depth, B D = 50 ft
7.3 FRONT WALL LOADS
The wall facing the explosion source is subjected to a reflection effect as the
blast wave impacts the facing wall and reflects back towards the blast source.
The reflection effect amplifies the blast pressure on the front or facing side of the
building. Because the overpressure at the top and side edges of the front wall is
FRONT WALL LOADS
vortex
rarefaction
wave
reflected
shock
front
rarefaction
wave
shock front
reflected
shock
front
187
shock front
vortex
Figure 7.2 Blast Wave at Front Wall (TNO Green Book)
less than the reflected overpressure (Figure 7.2), a decay in the reflection effect
takes place that starts at the edges and works inward. The effect is completely
removed after what is called the clearing time, tc . The clearing time is a function
of the height and width of the front wall.
The peak reflected overpressure, Pr , can be determined from Figure 7.3 using
the free-field peak overpressure. Alternatively, for free-field peak overpressures
Figure 7.3 Peak Incident Pressure versus the Ratio of Normal Reflected Pressure/
Incident Pressure for a Free-Air Burst (UFC 3-340-02)
188
BLAST LOADING
less than 40 psi and for sea level atmospheric pressures, the peak reflected overpressure can be determined using the following equation form Newmark:
Pr = [2 + 0.05 (Pso )]Pso
(7.2)
The clearing time can be calculated using the following equation from UFC 3340-02.
tc = 4S/[1 + S/G]Cr
(7.3)
In the preceding equation, S is the lesser of building height or building width.
G is the greater of building height or width, and Cr is the velocity of sound and
can be determined using Figure 7.4.
For the calculation of front wall dynamic wind pressure, a drag coefficient,
Cd , equal to 1.0 is used with the qo value determined from Figure 7.1.
To compute the remainder of the pressure-time curve, the following equations
are used for stagnation pressure, Ps , impulse, Is , and the effective duration, te ,
based on a simplified straight-line approximation. These values are illustrated in
Figure 7.5.
Ps = Pso + Cd (qo )
Is = 0.5 (Pr − Ps )tc + 0.5Ps to
te = 2Is /Pr
(7.4)
(7.5)
(7.6)
The blast wave angle of incidence affects the blast pressure load on the front
wall. This angle is taken as 0◦ for a blast wave traveling perpendicular into the
plane of the front wall where the full reflected overpressure is applied, and taken
as 90◦ for a blast wave traveling parallel to a surface, where the free-field, or sideon overpressure is applied. For intermediate values of the angle of incidence,
Figure 7.6 provides coefficients to calculate the applied pressure for use in the
following equation.
Pr α = (Cr α )(Pso )
(7.7)
7.3.1 Empirical Method—Front Wall Loading Example
The front wall is assumed to span vertically from foundation to roof. The calculation will be for a typical wall segment one foot wide.
With Pso = 10 psi, use Figure 7.3 to determine the reflected overpressure
ratio, Pr /Pso = 2.5.
Pr = (Pr /Pso )Pso = (2.5)(10 psi) = 25 psi
189
Figure 7.4 Velocity of Sound in Reflected Overpressure Region versus Peak Incident Overpressure (UFC 3-340-02)
190
BLAST LOADING
P
Pr
Equivalent Loading
Ps
tc
Figure 7.5
te
td
t
Front Wall Loading Diagram
Alternatively, use Equation 7.2 to compute the reflected overpressure,
Pr = [2 + 0.05 (Pso )]Pso = [2 + 0.05(10 psi)](10 psi) = 25 psi
To compute the reflected pressure clearing time, two parameters are needed,
S = minimum of B H or BW = minimum of 20 ft or 150 ft = 20 ft
G = larger of B H or BW = larger of 20 ft or 150 ft = 150 ft
With Pso = 10 psi, use Figure 7.4 to determine the sound velocity in the reflected region, Cr = 1.28 ft/ms
Use Equation 7.3 to compute the reflected overpressure clearing time,
tc = 4S/[1 + S/G]Cr = 4(20 ft)/[1 + (20 ft)/(150 ft)](1.28 ft/ms) = 55 ms
This result is nearly equal to to , thus tc ≈ to = 56 ms
Drag coefficient, Cd = 1.0
Use Equation 7.4 to calculate the stagnation pressure,
Ps = Pso + Cd (qo ) = (10 psi) + (1.0)(2.2 psi) = 12.2 psi
Use Equation 7.5 to calculate the front wall impulse,
Is = 0.5 (Pr − Ps )tc + 0.5Ps to = 0.5 [(25 psi) − (12.2 psi)](56 ms)
+ 0.5 (12.2 psi)(56 ms) = 700 psi-ms
Use Equation 7.6 to calculate the effective duration,
te = 2Is /Pr = 2 (700 psi-ms)/(25 psi) = 56 ms
pressure
25 psi
56 ms
time
191
Cr␣ = Pr␣/Pso
0
2
4
6
8
10
12
14
0
10
30
40
50
Angle of incidence, ␣ (Degrees)
60
70
Figure 7.6 Reflected Pressure Coefficient versus Angle of Incidence (UFC 3-340-02)
20
Peak Incident Overpressure, psi
5000
70
3000
50
2000
30
1000
20
500
10
400
5
300
2
200
1
150
0.5
100
0.2
80
90
192
BLAST LOADING
7.3.2 Empirical Method—Oblique Angle Example
Given angle of incidence, α = 40 degrees
Given side-on overpressure, Pso = 10 psi
Using Figure 7.6, determine the reflected pressure coefficient, Cr α = 2.4
Using Equation 7.7, calculate the reflected overpressure,
Pr α = (C1α )(Pso ) = (2.4)(10 psi) = 24 psi
7.4 SIDE WALL AND ROOF LOADS
Roofs and side walls represent surfaces that are parallel to the path of the advancing blast wave. There is no reflection effect for this situation; however, the
average pressure applicable to a specific area, for example a structural element,
depends on the length of the blast wave and the length of the structural element.
Figure 7.7 shows a diagram of a blast wave traveling across a roof. If a section of
the roof tributary to a supporting beam is oriented perpendicular to the traveling
wave, then the roof length would be short and the blast wave would have a full
effect. If the roof beam were oriented parallel to the traveling wave, then the roof
length would be longer, the average pressure would be reduced, and a rise time
would become evident. Calculations for these two situations are presented in the
side wall and roof examples. Values are obtained as follows:
r Equivalent uniform pressure factor, CE , is determined from Figure 7.8.
r Scaled uniform pressure rise time, tr , is determined from Figure 7.9.
r Scaled uniform pressure duration, to , is determined from Figure 7.10.
A building’s depth could be significant with respect to the distance from the
blast source. Thus the peak free-field overpressure for a side or rear surface could
be lower than that calculated for the front wall. The calculation of peak free-field
SHOCK FRONT
AT TIME t
f
b
Lw
L
Figure 7.7 Roof Loading Diagram
SIDE WALL AND ROOF LOADS
193
1
0.9
Positive Pressure, CE = PR/Psof
Negative Pressure, CE = PR/Psof
0.8
0.7
0.6
Equivalent Load Factor
0.5
0.4
0.3
Positive
0.2
Negative
0.1
0.2
0.3
0.4 0.5 0.60.70.8 1
2
3
Wave length/Span length, Lwf/L
4
5
6 7 8 9 10
Figure 7.8 Peak Equivalent Uniform Roof Pressure (UFC 3-340-02)
overpressure is determined using the criteria in Chapter 6. In many cases this
effect is negligible and is ignored, for simplicity.
For the calculation of roof and side wall dynamic wind pressure, a drag coefficient, Cd , determined from Table 7.1 is used with the qo value determined from
Figure 7.1.
194
BLAST LOADING
5
Numbers next to curves indicate
peak incident overpressure. psi
4
3
2
Scaled Rise Time
2
4
8
1
0.9
0.8
0.7
16
0.6
0.5
32
0.4
0.3
0.2
0.5
0.6
0.7
0.8
0.9
1
2
Wave length/Span length, Lwf/L
Figure 7.9 Scaled Rise Time of Equivalent Uniform Positive Roof Pressures (UFC 3340-02)
The total effective pressure on a roof or side wall is determined from the
following,
(7.8)
Pa = C E Pso + Cd qo
7.4.1 Empirical Method—Side Wall Loading Example
The side wall is the same as the front wall, spanning vertically from foundation
to roof. This calculation will be for a unit width wall segment, L = 1 ft.
SIDE WALL AND ROOF LOADS
195
Figure 7.10 Scaled Duration of Equivalent Uniform Roof Pressures (UFC 3-340-02)
196
BLAST LOADING
Table 7.1 Roof, Side Wall, and Rear Wall
Drag Coefficients
Peak dynamic pressure
Drag coefficient
0–25 psi
25–50 psi
50–130 psi
−0.40
−0.30
−0.20
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is needed,
L w /L = (78 ft)/(1 ft) = 78
With such a short wall segment in comparison to the blast wave length, it isn’t
necessary to obtain values from Figures 7.8, 7.9, and 7.10.
By inspection, C E = 1.0, use Equation 7.8 to calculate the effective peak
overpressure,
Pa = C E Pso + Cd qo = (1.0)(10 psi) + (−0.4)(2.2 psi) = 9.1 psi
By inspection, the effective rise time, tr = 0.0 ms
By inspection, the effective duration, td = 56 ms
pressure
9.1 psi
56 ms
time
7.4.2 Empirical Method—Roof Loading Example
The roof is a slab spanning between roof beams, and an appropriate span should
be used for L. However, to illustrate the use of the figures, a section spanning the
entire building will be used, thus L = 50 ft
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is needed,
L w /L = (78 ft)/(50 ft) = 1.6
With L w /L = 1.6, use Figure 7.8 to determine the effective uniform pressure
factor, C E ≈ 0.60
REAR WALL LOADS
197
Use Equation 7.8 to calculate the effective peak overpressure,
Pa = C E Pso + Cd qo = (0.60)(10 psi) + (−0.4)(2.2 psi) = 5.1 psi
With L w /L = 1.6, use Figure 7.9 to determine the effective scaled rise time,
tr /W 1/3 ≈ 1.4 ms/lb1/3
Rise time, tr = (tr /W 1/3 )(W 1/3 ) = (1.4 ms/lb1/3 )(10,000 lb)1/3 = 30 ms
With L w /L = 1.6, use Figure 7.10 to determine the effective scaled duration,
td /W 1/3 ≈ 3.0
Duration, td = (td /W 1/3 (W 1/3 ) = (3.0 ms/lb1/3 )(10,000 lb)1/3 = 65 ms
Total positive phase duration
to = tr + td = (30 ms) + (65 ms) = 95 ms
7.5 REAR WALL LOADS
The rear wall is the wall facing directly away from the blast source, as illustrated
in Figure 7.11. The calculation of rear wall blast loads is similar to that for the
side walls or roof. In fact, UFC 3-340-02 treats the rear wall as a roof extension.
For the calculation of rear wall dynamic wind pressure, a drag coefficient, Cd ,
determined from Table 7.1, is used with the qo value determined from Figure 7.1.
7.5.1 Empirical Method—Rear Wall Loading Example
The example will be for a wall segment 1 foot wide, L = building height = 20 ft
With qo = 2.2 psi, use Table 7.1 to determine the drag coefficient, Cd = −0.4
To determine the equivalent load values, the following parameter is
needed,
L w /L = (78 ft)/(20 ft) = 3.9
shock front
shock front
vortex
diffracted
shock
front
diffracted
shock
front
vortex
Figure 7.11 Rear Wall (TNO Green Book)
198
BLAST LOADING
With L w /L = 3.9, use Figure 7.8 to determine the effective uniform pressure
factor, C E ≈ 0.85
Use Equation 7.8 to calculate the effective peak overpressure,
Pa = C E Pso + Cd qo = (0.85)(10 psi) + (−0.4)(2.2 psi) = 7.6 psi
With L w /L = 3.9, extrapolate from Figure 7.9 to approximate the effective
scaled rise time, tr /W 1/3 ≈ 0.5 ms/lb1/3
rise time, tr = (tr /W 1/3 )(W 1/3 ) = (0.5 ms/lb1/3 )(10,000 lb)1/3 = 11 ms
Note that an alternative rise time calculation, based on the time for the shock
front to travel across the element, would be tr = L/Us = (20 ft)/(1.4 ft/ms) =
14 ms
With L w /L = 3.9, extrapolate from Figure 7.10 to approximate the effective
scaled duration, td /W 1/3 ≈ 2.1
Duration, td = (td /W 1/3 (W 1/3 ) = (2.1 ms/lb1/3 )(10,000 lb)1/3 = 45 ms
Note that the duration approaches the free field duration as L w /L gets large.
Total positive phase duration,
to = tr + td = (11 ms) + (45 ms) = 56 ms
pressure
7.6 psi
11 ms
56 ms
time
7.6 CONFINED EXPLOSIONS
As discussed in Chapter 6 an interior explosion involves more phenomena than
an external explosion, and the loading (Figure 7.12) changes accordingly. In the
internal explosion, there is first the directly incident shock as in an external explosion. This is followed by multiple shock reflections off the other surfaces
that are confining the explosion. Finally, there is a longer duration gas pressure
throughout the interior as the gaseous products of the detonation come to thermodynamic equilibrium.
In practice, it is often convenient to approximate this typical loading with a
suddenly applied bilinear pulse, as shown in Figure 7.13. The first part of this
loading is the directly incident shock as computed in Section 7.3 for front wall
loads. The second part is a triangular approximation to the gas pressure, which
generally has a lower peak but longer duration than the shock loading.
CONFINED EXPLOSIONS
199
Figure 7.12 Loading from Confined Explosion
To calculate this gas pressure loading, the peak gas pressure, Pg , is obtained from Figure 6.7, as described in the previous chapter. Then the scaled
gas impulse, i g /W 1/3 , is computed from Figures 7.14 through 7.25 as a function of loading density, W/V f , the scaled reflected shock impulse, ir /W 1/3 , and
the scaled mass, W F /W 1/3 , of the frangible openings in the confined volume.
Table 7.2 indicates which of the Figures 7.14 through 7.25 pertains to various
IDEALIZED SHOCK
PRESSURES
PRESSURE
ir = REFLECTED IMPULSE
IDEALIZED GAS
PRESSURES
TIME
Figure 7.13 Approximate Loading from Confined Explosion
200
BLAST LOADING
Table 7.2 Figures for Scaled Gas Impulse ig/W 1/3
W/Vf
W/Vf
W/Vf
W/Vf
= 0.002
= 0.015
= 0.15
= 1.0
ir /W 1/3 = 20
ir /W 1/3 = 100
ir /W 1/3 = 600
ir /W 1/3 = 2,000
Fig. 7.14
Fig. 7.17
Fig. 7.20
Fig. 7.15
Fig. 7.18
Fig. 7.21
Fig. 7.23
Fig. 7.16
Fig. 7.19
Fig. 7.22
Fig. 7.24
Fig. 7.25
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
2/3
A/Vf
Figure 7.14 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 20
10
201
CONFINED EXPLOSIONS
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
10
2/3
A/Vf
Figure 7.15 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 100
values of W/V f and ir /W 1/3 . For values between those given in this table, the
scaled gas impulse may be found by interpolation. Once the scaled gas impulse
is obtained, the actual impulse follows by unscaling, i.e., multiplication by W 1/3 .
The equivalent triangular duration of the gas loading is then tg = 2ir /Pg .
For example, let us estimate the loading from the detonation of W = 16 lbs
of TNT at the center of the floor a cubicle that measures 20 ft on a side. Suppose
202
BLAST LOADING
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
10000
1000
wr /w1/3
100
30
10
100
3
1
10
0.3
0
1
0.01
0.1
SCALED VENT AREA,
1
10
A/Vf2/3
Figure 7.16 Scaled Gas Impulse for W/V f = 0.002 and ir /W 1/3 = 600
that the total vent area is A = 20 ft2 and the weight of the frangible surfaces is
W F = 25 lbs.
Using the method described in Section 7.3, the directly incident shock is described by
Pr = 317 psi
to = 0.868 msec
203
CONFINED EXPLOSIONS
10000
SCALED GAS IMPULSE, ig/W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
1
0.1
SCALED VENT AREA,
A/Vf2/3
Figure 7.17 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 20
For the gas pressure, the loading density is
W/V f = 16/203 = 0.002 lbs/ft3
From Figure 6.7, the corresponding peak pressure is
Pg = 25.5 psi
10
204
BLAST LOADING
10000
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
1
100
0.3
10
0
1
0.01
0.1
SCALED VENT AREA,
1
10
A/Vf2/3
Figure 7.18 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 100
For the scaled reflected shock impulse, ir /W 1/3 = 54.6 psi × msec/lb1/3
Interpolation between Figures 7.14 and 7.15 gives the scaled gas impulse
ig /W 1/3 = 1990 psi × msec/lb1/3
The actual gas impulse is then
i g = 1990 × 161/3 = 5010 psi × msec
205
CONFINED EXPLOSIONS
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
10000
wr /w1/3
1000
100
30
10
3
100
1
0.3
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.19 Scaled Gas Impulse for W/V f = 0.015 and ir /W 1/3 = 600
and the equivalent triangular duration is
tg = 2 × 5010/25.5 = 393 msec
This dynamic loading is summarized in Figure 7.26. As to tg in this case,
the shock portion appears very close to the vertical axis.
206
BLAST LOADING
10000
wr /w1/3
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.20 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 20
7.7 LEAKAGE
If the openings in the exterior of a structure are not designed for an expected
blast loading, there will be a leakage of pressure into the interior. Idealized
pressure-time series for this loading are discussed in UFC 3-340-02 as a function
207
LEAKAGE
10000
wr /w1/3
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100
1000
30
10
3
1
0.3
100
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.21 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 100
of the structural geometry and shock wave characteristics. Different curves are
developed for the exterior front wall, interior front wall, interior side wall and
roof, and interior back wall. This manual also discusses closures that are deliberately designed to resist the blast but have small openings for vents and ducts
that admit some overpressure.
208
BLAST LOADING
10000
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
wr /w1/3
100
1000
30
10
3
100
1
0.3
10
0
1
0.01
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.22 Scaled Gas Impulse for W/V f = 0.15 and ir /W 1/3 = 600
7.8 RAY-TRACING PROCEDURES
Ray-tracing algorithms have been developed to provide a higher resolution of
the blast loading than provided by the previously discussed empirical methods in complex geometrical situations. These procedures generally work backwards from each location of interest on a structure to capture a finite number
209
RAY-TRACING PROCEDURES
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
30
1000
10
3
1
0.3
100
10
0.01
0
0.1
SCALED VENT AREA,
1
10
2/3
A/Vf
Figure 7.23 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 100
of the possible multiple reflections at each location. As more reflections are
considered, the fidelity increases, but so does the computational effort. In practice, the coefficients of such physics-based models are “tuned” to match test
data. Accordingly, these are often termed “semi-empirical” models. Their applicability is limited to configurations and charge weight ratios for which data are
available. Some semi-empirical codes have the capability to approximate shock
210
BLAST LOADING
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
30
1000
10
3
1
100
0.3
10
0.01
0
0.1
1
10
2/3
SCALED VENT AREA, A/Vf
Figure 7.24 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 600
diffraction, shielding, and reflection. In both fidelity and computational effort,
these methods lie between the previously discussed fully empirical procedures
and fully physics-based computational fluid dynamics. These semi-empirical
methods have been developed primarily by defense-related agencies and are restricted in distribution to the government and its contractors.
211
RAY-TRACING PROCEDURES
SCALED GAS IMPULSE, ig /W1/3 (psi - msec/lb1/3)
100000
10000
wr /w1/3
100
1000
30
10
3
100
1
0.3
10
0.01
0
0.1
1
2/3
SCALED VENT AREA, A/Vf
Figure 7.25 Scaled Gas Impulse for W/V f = 1.0 and ir /W 1/3 = 2,000
10
212
BLAST LOADING
Figure 7.26 Approximate Loading for Detonation of 16 lbs of TNT in a 20-ft Cubicle
7.9 SUMMARY
In this chapter, methods for determining the dynamic blast loading on structural
elements from the characteristics of blast phenomena have been discussed and
illustrated with examples. Emphasis was placed on simple approximate procedures that apply widely in practical blast-resistant design. More comprehensive
methods were also discussed.
REFERENCES
ASCE. 1985. Design of Structures to Resist Nuclear Weapons Effects (Manual No. 42).
Reston, VA: American Society of Civil Engineers.
ASCE Petrochemical. 1997. Design of Blast Resistant Buildings in Petrochemical Facilities. Reston, VA: Task Committee on Blast Resistant Design, American Society of
Civil Engineers.
ASCE Physical Security. 1999. Structural Design for Physical Security, Reston, VA:
American Society of Civil Engineers.
Biggs J. M. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill Book
Company.
Committee for the Prevention of Disasters Due to Dangerous Substances. 1992. Method
for the determination of possible damage to people and objects resulting from releases
of hazardous materials (CPR 16E) TNO Green Book. The Hague, The Netherlands:
The Director-General of Labour.
REFERENCES
213
Mays, G. C. and P. D. Smith, eds. 1995. Blast Effects on Buildings. London: Thomas
Tedford Publications.
Newmark, Nathan M. 1956. An engineering approach to blast-resistant design (Paper
2786). ASCE Transactions 121: 45–64. Reston, VA: American Society of Civil Engineers.
U.S. Department of Defense. 1965. Design of Structures to Resist the Effects of Atomic
Weapons (TM 5-856). Washington, DC: U.S. Department of Defense.
. 1990. Structures to Resist the Effects of Accidental Explosions (TM 5-1300).
Washington, DC: U.S. Department of Defense.
. 2008. Structures to Resist the Effects of Accidental Explosions
(UFC 3-340-02). Washington, DC: U.S. Department of Defense.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
8
Fragmentation
Kim King
8.1 INTRODUCTION
Injury and structural damage during an explosion may result not only from the
direct pressure and impulse of an air blast, but also from the impact of flying
objects called debris and fragmentation. Military weapons are typically explosive
charges with some type of metal casing. Upon detonation, this case is ruptured
and expelled as fragmentation at high velocity. Similarly, terrorist devices may be
embedded with objects such as ball bearings or nails that will be ejected at high
velocity. Other fragmentation may occur as objects interact with the blast wave
created from a detonation and become airborne. Fragments can range in size
from small irregularly shaped objects (fractions of an inch) to large objects such
as portions of buildings or vehicles. Two types of fragments may be generated
during an explosion: primary and secondary fragments.
Fragmentation generated during an explosion can damage structures by penetrating walls as well as impacting structural elements with sufficient impulse to
cause material failure or global structural failure of a structural element. This
chapter discusses fragmentation resulting from an explosion and established
methods of predicting the hazards of airborne fragments.
8.2 DEBRIS
Debris generically refers to the broken remains of a destroyed object. Debris
from a destroyed structure near an explosion can be thrown as secondary fragmentation.
8.3 LOADINGS
Fragmentation can typically be separated into two categories: primary and secondary fragmentation. Primary fragments are typically produced from the container or casing of an explosive charge. Primary fragments can also be generated in manufacturing processes where equipment or tools are in contact with
explosive material. Examples of primary fragments include the casing of
215
216
FRAGMENTATION
conventional munitions, metal containers such as kettles and hoppers used in
the manufacture of explosives, the steel casing of pipe bombs, and the metal
housing of rocket engines. For high explosives, the casing will typically shatter
into many small fragments traveling at very high initial velocities up to several
thousand feet per second. Primary fragments are typically small, on the order of
1 gram. They typically have “chunky” or irregular geometry with linear dimensions all of the same order. However, some casings are scored or formed to fail
in a specific manner to generate a specific fragment pattern.
Secondary fragments are produced when the shock wave encounters objects
or structures located near the source of detonation. Secondary fragments are also
known as “secondary debris” or “secondary missiles.” Secondary fragments are
generally larger than primary fragments and have lower initial velocities. Examples of secondary fragments include broken glass from window systems, debris from exterior walls, and dislodged architectural ornamentation. Secondary
fragmentation can also include items located near (but not in contact with) the
explosive source, such as parts of an automobile near a terrorist vehicle bomb.
Damage from secondary fragments is not easy to accurately predict because of
the variation in size, shape, and initial velocity.
8.3.1 Primary Fragmentation
Much work has been done to characterize the type of fragments and corresponding fragment velocity of military munitions. Many factors can affect the primary
fragment shape, mass, and velocity, including the explosive type, shape of the explosive charge, arrangement of casing, and so forth. For simplification, the most
common shape of a cased explosive (a cylinder) with uniform case thickness will
be considered.
Primary Fragment Shape The shape of primary fragments may be predictable
if, for example, the casing is scored to produce a particular shape, or a specific
munition, with a known fragment shape, is known. But a standard shape for
primary fragments is typically assumed for design purposes where a specific
shape is not known. The typical fragment shape is shown in Figure 8.1.
Primary Fragment Mass For the most common device shape, a uniform cylinder, and a given confidence level (CL ≤ 0.999), the design fragment weight can
be estimated with the following procedure:
Wd f = M A2 · ln2 (1 − C L )
where: Wdf = design fragment weight (ounces)
MA = fragment distribution factor
CL = confidence level
(8.1)
217
LOADINGS
d = Diameter of cylindrical portion of
fragment
r
r
d
r = Radius of hemispherical portion
of fragment (d/2)
r
r
r
d
Wf = Fragment weight, 0.645.d 3/γ = 0.186.d 3
D = Caliber density = 0.186.lb/in3
N = Nose shape factor = 0.72 + 0.25.(n – 0.25)1/2 = 0.845
n = Caliber radius of the tangent ogive of the fragment nose = r/d = 0.5
Figure 8.1 Standard Primary Fragment Shape
1/3
M A = B · tc5/6 · di
· (1 + tc di )
(8.2)
where: B = explosive constant (see Table 8.1)
tc = average casing thickness (inch)
di = average inside diameter of casing (inch)
Table 8.1 Mott Scaling Constants for Mild Steel Casings and Various
Explosives (Mott 1947)
A (oz1/2 inches−3/2 )
B (oz1/2 inches−7/6 )
Baratol
—
0.512
Composition A-3
—
0.22
Composition B
0.214
0.222
0.197
Explosive
Cyclotol (75/25)
—
H-6
—
0.276
HBX-1
—
0.256
HBX-3
—
0.323
Pentolite (50/50)
0.238
0.248
PTX-1
—
0.222
PTX-2
—
0.227
RDX
0.205
0.212
Tetryl
0.265
0.272
TNT
0.302
0.312
218
FRAGMENTATION
Table 8.2 Gurney Energy
Specific Weight (lb/inch3 )
Explosive
√
2 · E
Composition B
0.0621
9100
Composition C-3
0.0578
8800
HMX
0.0682
9750
Nitromethane
0.0411
7900
PBX-9404
0.0664
9500
PETN
0.0635
9600
RDX
0.0639
9600
TACOT
0.0581
7000
Tetryl
0.0585
8200
TNT
0.0588
8000
Trimonite No. 1
0.0397
3400
Tritonal (TNT/AL - 80/20)
0.0621
7600
Primary Fragmentation Velocity The most common technique for calculating
the initial velocity of primary fragmentation in contact with an explosive charge,
the Gurney method, is provided here (Gurney 1947). The initial velocity of primary fragments produced from a cylindrical cased explosive is a function of the
explosive output and the ratio of the explosive charge weight to casing weight.
The initial velocity of primary fragments resulting from a cylindrical cased explosive with evenly distributed explosives and casing is expressed as:
Vop
√
= FS · 2 · E ·
W
Wc
1 + 0.5 · W Wc
(8.3)
where: √ Vop = initial velocity of primary fragment (feet/sec)
2 · E = Gurney Energy Constant (see Table 8.2)
FS = factor of safety (1.2)
W = design charge weight = FS·Wact (feet/sec)
Wact = actual quantity of explosive material (lb)
Wc = weight of casing (lb)
8.3.2 Secondary Fragmentation
Secondary fragments generated during an explosion can represent a wide variety of sizes, shapes, and initial velocities. An array of scenarios must typically
be considered to establish the design fragment (worst-case loading on the structure). Varying factors such as fragment size, shape, structural constraint, and
orientation must be evaluated to establish the controlling fragment scenario that
produces the most severe response of the structure in question.
LOADINGS
219
Pressure
Pr
CD.Q
F (t) = CD.q(t) = CD.1/2.ρu 2
ta
t1
t2
t3
Time
Figure 8.2 Idealized Time Loading on an Irregular Fragment
Initial steps for identifying the hazard associated with secondary fragmentation are listed below.
1. Define blast loading (see Chapter 7 for blast loading).
2. Determine distance from the center of the explosive to the potential fragment (refer to structural details of the equipment and/or structure in consideration).
3. Determine the size, shape, and connection of the potential fragment (see
structural details for the equipment and/or structure in consideration).
4. Determine potential fragment velocity using the methods detailed in this
chapter.
The blast load applied to the object can be simplified to a three-step function
that represents the rise to peak pressure, initial pressure decay that is affected by
diffraction of the wave around the object and rarefaction waves that form across
the front of the object, and finally the drag phase loading (see Figure 8.2).
Unconstrained Fragments Unconstrained fragments are generated when objects that are not attached interact with the blast wave and are thrown from the
incident origin. Predictions for unconstrained fragment velocity assume that the
object is not attached (i.e., no energy lost breaking the object free), the object
will not deform, and gravity will not affect velocity during acceleration.
220
FRAGMENTATION
Unconstrained Fragments Far from the Charge Potential fragments are defined
as far from the charge if the objects are located at a distance more than 20 times
the radius of the explosive. For potential fragments located far from the charge,
the initial fragment velocity can be calculated using the following methods (U.S.
Department of Defense 2008).
The basic equation of motion describes the object acceleration, where the
time-dependent pressure applied to the area of the object facing the blast is equal
to the mass of the object times acceleration. This relationship can be integrated
and rearranged to yield:
td
V (td ) =
A
id
M
(8.4)
0
where: V = velocity (feet/sec)
td = load duration (msec)
A = area of object facing blast (inches2 )
M = mass of object (lb-msec2 /inch)
id = total drag and diffraction (psi-msec/lb1/3 )
Equation 8.4 can be explicitly resolved if the loading history can be mathematically defined in a function. However, Figure 8.3 can be used to graphically
solve for the nondimensional velocity and then solve for the object initial velocity, by solving for the quantity Pso /Po (on the ordinate) and the following value
on the abscissa:
12 · C D · i s · ao
1000 · Pso · (K · H + X )
(8.5)
Establish the nondimensional object velocity using the various curves. The
nondimensional velocity (¯v ) is equal to:
v¯ =
144 · vo · M · ao
106 · Po · (K · H + X )
(8.6)
Rearranging Equation 8.6 will result in the initial object velocity.
where: Pso = peak incident overpressure (psi)
Po = atmospheric pressure (psi)
Cd = drag coefficient (see Figure 8.4)
is = incident-specific impulse (psi-msec)
ao = velocity of sound in air (feet/sec)
K = constant (see Table 8.3)
H = minimum transverse dimension of the mean presented area of
object (inch)
LOADINGS
221
5
1
PsoPo
0.1
50.0
0.01
10.0
5.0
1.0
0.5
0.001
0.0001
2E-5
0.2
0.1
0.05
Numbers adjacent to curves indicate
nondimensional object velocity =
(144 vo.M.ao)/106.Po.A.(K.H+X)
1
10
0.01
0.005
0.001
100
1,000
10,000
100,000
1,000,000
(12.Cd.js.ao)/(1000.Pso.(K.H +X))
Figure 8.3 Nondimensional Object Velocity
X = distance from the front of the object to the location of its largest
cross section normal to the plane of the shock front (inch)
vo = initial velocity of object (feet/sec)
Unconstrained Fragments Near the Charge Potential fragments are defined as
near the charge if the objects are located at a distance less than or equal to 20
times the radius of explosive. Initial fragment velocity for potential fragments
can be calculated using the following methods (U.S. Department of Defense
2008).
The first step to determine the velocity of a close-in unconstrained fragment
is to determine the specific impulse. Two charge configurations are considered,
spherical and cylindrical. For spherical charges with R/Re ≤ 5.07:
i
=
β · Reff
Re
Rt
0.158
Re
= 38,000 ·
R
1.4
(8.7)
and Reff = Re
For cylindrical charges with R/Re ≤ 5.25:
i
=
β · Reff
Re
Rt
0.158
Re
= 46,500 ·
R
(8.8)
222
FRAGMENTATION
Shape
Sketch
CO
FLOW
Circular cylinder (Long rod),
side-on
1.20
FLOW
0.47
Sphere
FLOW
0.82
Rod, end-on
FLOW
OR
Disc, face-on
FLOW
Cube, face-on
Cube, edge-on
FLOW
Long rectangular member,
face-on
FLOW
Long rectangular member,
edge-on
FLOW
1.17
1.05
0.80
2.05
1.55
FLOW
Narrow strip, face-on
Figure 8.4 Drag Coefficients (Baker et al. 1977)
1.98
LOADINGS
223
Table 8.3 K Constant
Steel
K-Value
Object on reflecting surface (i.e., on ground)
4
Object in air
2
For cylindrical charges with 5.25 < R/Re ≤ 10:
i
=
β · Reff
Re
Rt
0.158
Re
= 161,700 ·
R
1.75
(8.9)
and
Reff
where:
Le
= 0.909 ·
Re
0.333
· Re
(8.10)
i = specific acquired impulse (psi-msec)
β = nondimensional shape factor of the target from Figure 8.5
Reff = effective radius (feet)
Re = radius of the explosive (feet)
R = standoff distance (feet)
Rt = target radius (feet)
Le = length of cylindrical explosive (inch)
Figure 8.6 graphically shows specific acquired impulse for spheres and cylinders. The data for this relationship were experimentally derived and should not
be used beyond the distances shown in the figure. When these standoff distances
are greater than plotted values, the normal reflected impulse may be used as an
approximation for specific acquired impulse.
Because the target is so close to the charge, the actual load history is unimportant, and the impulse acting on the target is equal to the applied momentum.
Therefore, i=M·V/A and the velocity of a specific target is:
vo =
1000 · i · β · A
12 · M
(8.11)
Constrained Fragments A constrained secondary fragment is generated when
an object is torn loose from its moorings and thrown from the explosion source.
The velocity of a constrained secondary fragment is dependent on the specific
impulse applied to the object minus the impulse required to free the object from
the support. Therefore, the following is true:
I − Ist = M · vo
(8.12)
224
FRAGMENTATION
Cylindrical Explosive
2Re
2Rt
Le
a)
b)
Exposed Flat
Face β = 1.0
Exposed Cylindrical
Surface β = π/4
c)
Exposed Spherical
Surface β = 2/3
Spherical Explosive
2Re
2Rt
R
Figure 8.5 Nondimensional Shape Factor (U.S. Department of Defense 2008)
50,000
Specific Impulse
Cylindrical Charge
Spherical Charge
10,000
5,000
2,000
1
5
R/Re
Figure 8.6 Specific Acquired Impulse
10
LOADINGS
Table 8.4
225
Material Toughness
Toughness inches-lbf/inches3
Steel
ASTM A36
12,000
ASTM A441
15,000
ASTM A514 Grade F
19,000
where: I = total blast impulse applied to the fragment (psi-msec)
Ist = impulse required to free the fragment from the support (psi-msec)
Empirical methods have been produced to estimate the initial fragment velocity from a cantilever or a fixed-fixed beam (clamped-clamped) (U.S. Department
of Defense 2008). If the following relationship is true, the object velocity is zero
because the applied impulse is insufficient to disengage the object.
If:
i · bf
≤ 0.602,
2 · L b 0.3
ρ f · T · Ab ·
bf
where:
(8.13)
i = unit impulse applied to the object (psi-msec)
bf = width of fragment exposed to blast (inch)
ρ f = mass density of the fragment lb-msec2 /inch4 )
Ab = smallest cross-sectional area of the fragment (inch2 )
Lb = length of fragment exposed to blast (inch)
T = toughness of material (see Table 8.4)
Then, vo = 0 · ft s [m/s]; otherwise:
⎡
⎛
⎢
⎜
1000
⎢
⎜
T
·
·
+
K
·
vo =
K
⎢ 1
2 ⎜
ρf
⎣
⎝
12
⎞⎤
⎟⎥
i · bf
⎟⎥
0.3 ⎟⎥ (8.14)
⎠⎦
2 · Lb
ρ f · T · Ab ·
bf
where: K 1 = constant (see Table 8.5)
K 2 = constant (see Table 8.5)
Table 8.5
Constants K1 and K2
Support Condition
K1
K2
ASTM A36
−0.2369
+0.3931
ASTM A441
−0.6498
+0.4358
226
FRAGMENTATION
4
Cantilever
Clamped-Clamped
3
.
v2 = 12 V . ρf .T
1000 √
2
1
0
0
1.5
3
4.5
2.Lb
i .bf .
6
7.5
9
10.5
0.3
bf
Ab ρt .T
√
Figure 8.7 Scaled Fragment Velocities for Constrained Objects
Figure 8.7 shows Equation 8.14 graphically for both cantilever and fixed-fixed
support conditions.
8.4 DESIGN FRAGMENT PARAMETERS
As a fragment is thrown from its initial position, several factors influence the
impact the fragment will have on the target. Those factors include the fragment
trajectory, the velocity upon impact, and the angle at which the fragment actually
impacts the target (referred to as angle of obliquity). A zero angle of obliquity is
the most conservative and typically assumed for design, since the actual angle of
obliquity is rarely known.
8.4.1 Fragment Final Velocity
Fragment velocities vary with distance due primarily to drag forces. As the fragment travels away from the event, the initial velocity will decay until impact.
It is assumed the initial velocity is achieved instantaneously. If a target is located close to an explosive event (less than 20 feet [U.S. Department of Defense
2008]), the decay in velocity is minimal and can be ignored. For targets farther
from the event, the following method can estimate the final fragment velocity.
Vs = vo · e−[12·kv ·R f ]
(8.15)
DESIGN FRAGMENT PARAMETERS
227
where: Vs = fragment velocity at distance of interest (Rf )
Rf = distance from the center of detonation
kv = velocity decay coefficient
and:
kv =
where:
A
· ρa · C D
Wf
(8.16)
Wf = fragment weight (ounces)
A/Wf = fragment form factor
ρ a = specific density of air (ounces/inch3 )
For primary fragments, some basic assumptions simplify the equation for final
fragment velocity.
Assume:
1/3
A/W f = 0.78/W f (for mild steel fragments)
ρa = 0.00071−ounces/inch3
C D = 0.6
Resulting in:
Vs = vo · e
− 0.004·
Rf
W f 1/3
(8.17)
8.4.2 Fragment Trajectory
Once the initial velocity of primary or secondary fragments is determined, it
will be necessary to establish the potential range of the accelerated fragments, to
establish the hazard associated with the fragments. The range is established by
solving the equations that define the trajectory the fragment will follow until it
strikes the ground or a target. Because fragments generated during an explosion
are typically irregular in shape and mass, determining the fluid dynamic forces
acting on the fragments to influence the trajectory can be difficult, and simplified
fluid dynamic force prediction methods are typically adopted. Baker describes
the acceleration in the X and Y directions based on both lift and drag forces
(Baker et al. 1983). However, explosively produced primary and secondary fragments are typically “chunky” or drag-controlled fragments. Thus, most methods
ignore the lift forces and evaluate the accelerations based on drag forces only
(U.S. Department of Defense 2008).
The equations that define the accelerations and velocities in the X and Y directions can be solved simultaneously to determine the distance traveled by the
fragment. Figure 8.8 represents a series of calculations for the resulting fragment
range given a variety of initial conditions. A number of initial trajectories were
228
FRAGMENTATION
10
(12.ρ0.CD.ADR)/M
5
1
0.5
0.1
0.05
0.01
0.01
1
10
(12.ρ0.CD.ADv02)/(M.g)
0.1
100
1,000
Figure 8.8 Fragment Range Prediction
considered to determine the maximum range; thus, the initial trajectory does not
have to be established to use Figure 8.8. Calculate the nondimensional velocity
[v], shown on the abscissa in Figure 8.8, with the following:
v=
12 · ρo · Cd · A D · vo2
M·g
(8.18)
where: ρ o = mass density of the medium through which the object travels (lbmsec2 /inch4 )
AD = drag area (inch2 )
g = gravity force (32.2 feet/sec2 )
From Figure 8.8 determine the dimensional range [R], shown on the ordinate.
Rearrange R to solve for the maximum range [R].
R=
M·R
12 · ρo · Cd · A D
(8.19)
8.5 FRAGMENT IMPACT DAMAGE
Almost any type of structure is susceptible to damage from explosively driven
fragmentation. Damage can range from cosmetic damage such as bent or cracked
architectural features to perforation (where a fragment passes through a wall
or structural element), penetration (where a fragment strikes and lodges in a
FRAGMENT IMPACT DAMAGE
229
structural element), or even global collapse (when a larger fragment strikes a
target with enough momentum to cause a primary element to fail).
Detonation of explosives can result in the formation of primary and secondary
fragmentation. These fragments can range from small primary fragments with
initial velocities in the order of thousands of feet per second to larger secondary
fragments with initial velocities of hundreds of feet per second. When a fragment
strikes a target, it will perforate the target, become embedded in the target (i.e.,
penetrate the target either with or without spalling), or be completely deflected by
the target. Factors that influence the fragment/target response include the initial
fragment velocity, the distance between the explosion and the target, the angle
at which the fragment strikes the target (angle of obliquity), and the physical
properties of the fragment (mass, shape, and material strength) and the target
(strength and thickness).
8.5.1 Fragment Penetration into Miscellaneous Materials
(THOR Equation)
Empirical THOR equations were developed based on testing for metallic and
nonmetallic materials (Ballistic Analysis Laboratory 1961, Ballistic Research
Laboratory 1963). All testing was performed to simulate primary fragments, and
the resulting data should be applied accordingly. The test data were assembled
into the THOR reports by Greenspon in 1976 (Greenspon 1976). These data
were gathered for military purposes and as such should be used in commercial
and industrial structures with specific limitations in mind. The method assumes
small fragment sizes into specific targets. The equation for minimum thickness
of a plate to resist perforation by a mild steel fragment is:
t=
where:
10 −
C1 +7C2 +3C3
C2
Af
·mf
C
− C3
2
· Vs
1−C5
C2
C4
· (cos φ) C2
(8.20)
t = fragment penetration (m)
mf = fragment mass (kg)
Af = presented area of fragment (m2 )
Vs = impact velocity (m/sec)
φ = impact angle relative to normal from tangent (rad)
C1 – C10 = empirical constants (see Table 8.6)
Other useful THOR equations for application to fragment impact on structures
are residual velocity and residual fragment mass equations. The THOR equation
for residual velocity is:
C
Vr = Vs − 10(C1 +7C2 +3C3 ) · t · A f 2 · m Cf 3 · (1/ cos φ)C4 · VsC5
where: Vr = residual velocity of fragment after perforation (m/sec)
(8.21)
230
1.044
1.112
0.203
0.328
2.534
0.674
2.590
1.310
−0.093
1.631
2.877
2.801
2.118
0.441
Lead
Lexan
Magnesium
Nylon, bonded
Nylon, unbonded
Plexiglass, as cast
Plexiglass, stretched
Steel, face hardened
Steel, hard homogeneous
Steel, mild homogeneous
Titanium
Tuballoy
1.021
0.583
1.103
0.889
0.889
0.674
0.835
1.144
1.092
0.720
0.499
0.705
3.431
0.678
1.046
Copper
1.029
1.042
Glass, bullet-resistant
0.314
Cast iron
C2
Doron
2.907
1.037
Aluminum (2024-T3)
C1
Target Material
Table 8.6 THOR Constants
1.369
1.073
−1.035
0.865
0.990
−0.654
−1.095
0.743
−0.968
−0.603
1.050
−1.117
1.262
0.773
−1.657
−0.945
0.655
−0.502
1.262
0.690
−0.723
−0.945
0.917
−1.014
0.989
0.846
−0.730
0.715
1.028
−1.051
−0.791
1.250
−1.072
−0.903
C4
C3
0.828
0.167
0.019
0.019
0.434
0.686
0.242
−0.162
0.392
−0.087
0.603
0.818
0.465
−0.362
0.802
0.523
−0.139
C5
−4.564
−1.927
−2.616
−3.017
−1.768
−6.431
−6.115
−6.619
−12.165
−6.026
−7.063
−3.267
−6.341
−10.161
−5.904
−9.052
−6.548
C6
0.560
1.086
0.138
0.346
0.234
0.437
1.402
−0.067
0.035
0.285
0.480
0.506
0.305
0.251
0.340
0.162
0.227
C7
0.447
−0.748
0.835
0.629
0.744
0.169
−0.137
0.903
0.775
0.803
0.465
0.350
0.429
0.343
0.568
0.673
0.694
C8
0.640
1.327
0.143
0.327
0.469
0.620
0.674
−0.351
0.045
−0.172
1.171
0.777
0.747
0.706
1.422
2.091
−0.361
C9
1.381
0.459
0.761
0.880
0.483
1.683
1.324
1.717
3.451
1.519
1.765
0.934
1.819
2.906
1.650
2.710
1.901
C10
231
FRAGMENT IMPACT DAMAGE
The THOR equation for residual fragment weight is:
C
m r = m f − 10(C6 +7C7 +3C8 −3) · t · A f 7 · m Cf 8 · (1/ cos φ)C9 · VsC10
(8.22)
where: mr = residual mass (kg)
The range for each variable in the THOR equations is given in Table 8.7.
8.5.2 Steel
Limit Velocity of “Chunky” Nondeforming Fragments for a Thin Metal Target
Fragment penetration into thin metal structural elements such as typical metal
buildings can be calculated based on the metal target material and fragment properties. A fragment velocity greater than the limit velocity (V 50 ) will penetrate the
metal target. The limit velocity for a thin metal target can be calculated using:
√
σt · ρt · V50n
(8.23)
V50 =
ρf
where:
σ t = target yield stress
ρ t = target density
ρ f = fragment density
V 50n = nondimensional limit velocity (calculate h/a and determine V 50n
from Figure 8.9)
h = target thickness
a = fragment radius
Fragment Penetration into Mild Steel Targets Fragment penetration calculations are based on some simplifying assumptions. The primary assumption is
that the fragment shape is a normal cylinder with a rounded leading edge (see
Figure 8.1). Additionally, the steel material is assumed to have a Brinell hardness of less than 150. Results will be conservative for materials with increased
hardness. Depth of penetration for mild steel targets impacted by armor-piercing
steel fragments is:
· Vs1.22
x = 0.30 · W 0.33
f
(8.24)
Depth of penetration for mild steel targets impacted by mild steel fragments is:
x = 0.21 · W 0.33
· Vs1.22
f
where:
x = depth of penetration (inches)
Wf = fragment weight (ounces)
V s = striking velocity (feet/sec)
(8.25)
232
Target Thickness
Range t mm
0.5 to 51.0
4.8 to 14.0
1.5 to 25.0
1.3 to 38.0
5.0 to 42.0
1.8 to 25.0
3.2 to 25.0
1.3 to 76.0
11.0 to 51.0
0.5 to 76.0
5.0 to 28.0
1.3 to 25.0
3.6 to 13.0
8.0 to 25.0
1.0 to 30.0
2.5 to 5.0
Target Material
Aluminum (2024-T 3)
Cast iron
Copper
Doron
Glass, bullet-resistant
Lead
Lexan
Magnesium alloy
Nylon, bonded
Nylon, unbonded
Plexiglass, as cast
Plexiglass, stretched
Steel, face hardened
Steel, homogeneous
Titanium alloy
Tuballoy
Table 8.7 Range of Variables in THOR Equations
1,372 to 3,078
213.4 to 3,170
182.9 to 3,658
762 to 2,987
152.4 to 3,353
61.0 to 2,897
91.4 to 3,048
304.8 to 3,658
152.4 to 3,200
304.8 to 3,505
152.4 to 3,170
61.0 to 3,048
152.4 to 3,353
335.4 to 3,475
335.0 to 1,859
365.8 to 3,353
Striking Velocity
Range Vs m/sec
1.9 × 10−3 to 3. 10 × 10−2 (30 to 475)
1.9 × 10−3 to 1.60 × 10−2 (30 to 240)
3.2 × 10−4 to 5.30 × 10−2 (5 to 825)
9.7 × 10−4 to 1.60 × l0−2 (1 5 to 240)
4.6 × 10−8 to 4. 4 × 10−6 (7.14 × 10−4 to 6.79 × 10−2)
4. 6 × 10−8 to 4.4 × 10 −6 (7.14 × 10−4 to 6.79 ×
10−2 )
4 .6 × 10−8 to 1.9 × 10−6 (7.14 × 10−4 to 2.96 × 10−2 )
4 .6 × 10−8 to 7.6 × 10−6 (7.14 × l 0−4 to 1.18 × 10−1 )
9.7 × 10−4 to 1.60 × l0−2 (15 to 240)
4.6 × 10−8 to 2.2 × 10−6 (7.14 × 10−4 to 3.43 × 10−2 )
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
1.4 × 10−7 to 4.4 × 10−6 (2.14 × 10−3 to 6.79 × 10−2 )
2.3 × 10−8 to 5.6 × 10−6 (3.57 × 10−4 to 8.57 × 10−2 )
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
9.7 × 10−4 to 1.60 × 10−2 (15 to 240)
3.2 × 10−4 to 1.60 × 10−2 (5 to 240)
Fragment Size Range mf kg (grains)
FRAGMENT IMPACT DAMAGE
233
12
10
ρf .V50
√ σt . ρt
8
6
4
2
0
0.2
0.4
0.6
0.8
1.0
1.2
h/a
1.4
1.6
1.8
2.0
2.2
Figure 8.9 Nondimensional Limit Velocity versus Nondimensional Thickness for
“Chunky” Nondeforming Fragments (U.S. Department of Energy 1992)
Most fragments generated during an explosion will not resemble a simplified
bullet—the majority of primary fragments are more blunt; thus, the bullet-shaped
fragment is conservative. Certainly some sharper-edged fragments are generated;
however, the likelihood that these fragments will strike the target in the appropriate orientation to control the design is minimal.
8.5.3 Fragment Penetration into Concrete Targets
When a fragment strikes a concrete target, the impulse imparted by the fragment
on the concrete will cause a crater of irregular size. The depth of the crater grows
with impact velocity for a given fragment size (or with a reduction in fragment
cross-sectional area for a given velocity). If the velocity is high enough (typically
greater than 1,000 feet/second [U.S. Department of Defense 2008]), the fragment
will penetrate beyond the crater. If the fragment deforms upon impact, the crater
size will decrease or may not form at all.
In addition to the crater on the impact side, a crater may form (generating concrete spall) on the opposite side of the target adjacent to impact. This crater is the
result of a compression wave passing through the concrete material and reflecting
off the free surface of the opposite concrete face. The reflection causes tension
stresses at the surface of the concrete element. If the tension stress exceeds the
compressive strength of the concrete, spalling of the concrete target will occur.
If the impulse imparted on the target is sufficient, the spall crater may extend to
the reinforcing steel. As impact velocity increases, the impact and spall craters
increase, and the fragment will become lodged in the target or pass through the
target (perforate).
234
FRAGMENTATION
Armor-Piercing Fragment Penetration into Concrete Targets The maximum
penetration of an armor-piercing fragment into a massive concrete target is defined as:
For Xf ≤ 2·d:
X f = 4.0 · 10−3 ·
√
K · N · D · d 1.1 · Vs0.9
(8.26)
For Xf > 2·d:
X f = 4.0 · 10−6 · (K · N · D) · d 1.2 · Vs1.8 + d
(8.27)
where: Xf = maximum penetration by armor-piercing
fragment (inch)
K = penetration constant = K = 12.91
f c
N = nose shape factor (see Figure 8.1)
D = caliber density (see Figure 8.1)
d = fragment diameter (inch)
For a standard primary fragment and concrete strength (f c ) of 4,000 psi, Equations 8.26 and 8.27 can be expressed in terms of fragment diameter (inches) as:
For Xf ≤ 2·d:
X f = 2.86 · 10−3 · d 1.1 · Vs0.9
(8.28)
X f = 2.04 · 10−6 · d 1.2 · Vs1.8 + d
(8.29)
For Xf > 2·d:
Or in terms of fragment weight:
For Xf ≤ 2·d:
X f = 1.92 · 10−3 · W 0.37
· Vs0.9
f
(8.30)
1.8
+ 0.695 · W 0.33
X f = 1.32 · 10−6 · W 0.4
f · Vs
f
(8.31)
For Xf > 2·d:
For concrete strengths other than 4,000 psi, the penetration depth can be estimated by multiplying the results by the square root of the ratio of the concrete
strengths:
X f = X f ·
4000
f c
(8.32)
where: X f = maximum penetration by armor-piercing fragment into concrete
with compressive strength of f c (inch)
FRAGMENT IMPACT DAMAGE
Table 8.8
235
Fragment Penetration Factors, K3
Type of Material
K3
Armorpiercing steel
1.00
Mild steel
0.70
Lead
0.50
Aluminum
0.15
Non-Armor-Piercing Fragment Penetration into Concrete Targets For nonarmor-piercing fragments, the depth of penetration into a concrete target can be
calculated by taking the fragment metal hardness into account:
X f = K3 · X f
(8.33)
where: X f = maximum penetration by non-armor-piercing fragment (inch)
K 3 = constant to account for fragment metal hardness (see Table 8.8)
8.5.4 Fragment Perforation of Concrete Targets
The equations presented in Section 8.5.3 assume the concrete is of infinite thickness. Of course the material thickness of a concrete slab of finite thickness required to prevent perforation is greater than the maximum penetration depth
into an infinite thickness (due to confinement and compressive resistance of a
massive concrete structure). The minimum concrete thickness required to prevent perforation as a function of maximum penetration into a massive concrete
structure is:
T p f = 1.13 · X f · d 0.1 + 1.311 · d
(8.34)
where: Tpf = minimum concrete thickness to prevent perforation by design fragment (inch)
Xf = maximum fragment penetration into a massive concrete structure
as described in Section 8.5.3
If a fragment perforates a concrete target, it can pose a threat to personnel
and structural elements behind the concrete target. The residual velocity of the
fragment can be calculated using equations that define the fragment velocity as
a function of penetration depth:
For Xf ≤ 2·d:
0.555
Tc 2
Vr
= 1−·
Vs
Tp f
(8.35)
236
FRAGMENTATION
For Xf > 2·d:
0.555
Tc
Vr
= 1−·
Vs
Tp f
(8.36)
where: Tc = concrete thickness up to Tpf (inch)
Vr = residual velocity of fragment as it leaves the element (feet/sec)
8.5.5 Fragment Spalling of Concrete Targets
As discussed in Section 8.5.3, concrete spall occurs when the impacting fragment
imparts enough impulse to generate a compression wave and a resulting tension
wave in the concrete that exceeds the compressive strength of the concrete. The
spall crater typically does not exceed the reinforcement depth. Concrete spalling
is a function of fragment penetration depth. Therefore, the minimum concrete
thickness required to prevent spalling as a function of maximum penetration into
a massive concrete structure can be expressed as:
Tsp = 1.215 · X f · d 0.1 + 2.12 · d
(8.37)
where: Tsp = minimum concrete thickness to prevent spall by design fragment
(inch)
Xf = maximum fragment penetration into a massive concrete structure
as described in Section 8.5.3 (inch)
If spalling occurs, it is likely the concrete spall will generate a secondary fragment that is hazardous to personnel and equipment behind the concrete target.
While the velocity produced when the concrete material fails is likely to be relatively low, the spall fragment velocity will also be affected by the wall motion
in response to the blast loading. Therefore, it is typically accepted that concrete
spall should be prevented by designing the concrete thickness to be sufficient to
avoid spall, or by providing an additional cover for the concrete surface (i.e., a
catch system or fiber-reinforced laminate).
8.5.6 Roofing Materials
Most roofing materials are relatively lightweight and have low thresholds of serious damage. However, the angle of obliquity will likely reduce the effect of fragment damage. Lower limits for fragment damage to miscellaneous lightweight
roofing materials based on fragment momentum are shown in Table 8.9.
REFERENCES
237
Table 8.9 Fragment Impact Damage for Roofing Materials (Baker et al. 1977)
Roofing Material
Minimum Fragment
Momentum (mV) for
Serious Damage
(lbf-sec)
Comments
Asphalt shingles
0.159
CracksShingle
1.370
Damage deck
<0.159
Crack tar flood coat
0.451
Crack surface of conventional built-up
roof (BUR) without top layer of stones
>0.991
Crack surface of conventional BUR with
2.867 lb/feet2 (137 Pa) top layer of
stones
Built-up roof
Miscellaneous
1/8-inch asbestos cement
shingles
0.159
1/4-inch asbestos cement
shingles
0.285
1/4-inch green slate
0.285
1/4-inch gray slate
0.159
1.2-inch cedar shingles
0.159
3/4-inch red clay tile
0.285
Standing seam terne metal
0.991
Plywood deck cracked
8.5.7 Other Materials
A number of other materials have been evaluated for fragment penetration, including sand, various rock and other geological materials, and combinations of
sand and concrete panels. While they are not presented here, they are covered in
varying detail in Structures to Resist the Effects of Accidental Explosions (U.S.
Department of Defense 2008), Explosion Hazards and Evaluation (Baker et al.
1983), and A Manual for the Prediction of Blast and Fragment Loadings on
Structures (Department of Energy 1992), as well as the Department of the Army
Historical Summary (DAHS) manual Fundamentals of Protective Structure Design for Conventional Weapons (Department of the Army 1998).
REFERENCES
Baker W. E., P. A. Cox, P. S. Westine, J. J. Kulesz, and R. A. Strehlow. 1983. Explosion
Hazards and Evaluation. Fundamental Studies in Engineering 5. Amsterdam: Elsevier
Scientific Publishing Company.
Baker W. E., J. J. Kulesz, R. E. Ricker, R. L. Bessey, P. S. Westine, v. b. Parr, and g. a.
Oldham. 1977. Workbook for Predicting Pressure Wave and Fragmentation Effects
238
FRAGMENTATION
of Exploding Propellant Tanks and Gas Storage Vessels, NASA CR-134906, NASA
Lewis Research Center.
Ballistic Analysis Laboratory, Johns Hopkins University. 1961. The Resistance of Various Metallic Materials to Perforation by Steel Fragments; Empirical Relationships for
Fragments Residual Velocity and Residual Weight (Project THOR Technical Report
47). Baltimore, MD: Johns Hopkins University, Institute for Cooperative Research.
Ballistic Research Laboratory. 1963. The Resistance of Various Non-Metallic Materials to Perforation by Steel Fragments, Empirical Relationships for Fragment Residual
Velocity and Residual Weight (TR 51). Aberdeen Proving Ground, MD: Ballistic Research Laboratory.
Department of the Army. 1998. Fundamentals of Protective Structure Design for Conventional Weapons (TM 5-855-1, Air Force AFPAM 32-1147[I], Navy P-1080, DSWA
DAHSCWEMAN-97). For Official Use Only. Washington, DC: Department of the
Army.
Greenspon J. E. 1976. An Approximate Non-Dimensional Representation of the THOR
Equations (U.S. Army Materiel Systems Analysis Activity TR #173).
Gurney R. W. 1947. The Initial Velocities of Fragments from Bombs, Shells, and Grenades
(Report No. 648). Aberdeen Proving Ground, MD: Ballistic Research Laboratory.
Mott R. I. 1943. A Theoretical Formula for the Distribution of Weights of Fragments
(Ministry of Supply, AC-3642).
U.S. Department of Defense (DoD). 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense.
U.S. Department of Energy. 1992. A Manual for the Prediction of Blast and Fragment
Loadings on Structures (DOE/TIC-11268). San Antonio, TX: Southwest Research Institute.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
III
System Analysis and Design
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
9
Structural Systems Design
Robert Smilowitz and Darren Tennant
9.1 GENERAL DISCUSSION
In order to protect the occupants of a building that is either the target of an explosive attack or the unintended victim of collateral damage resulting from an attack
against a neighboring property, attention must be paid to the performance of the
structural system. Although the structure may require extensive repair following
an incident, the primary goal of protective design is to prevent structural collapse
and to minimize debris. This goal is accomplished through the design and detailing of structural components to sustain specified intensities of blast loading, and
the design and detailing of structural systems to provide ductile and redundant
load paths, regardless of the cause or event. This combination of threat-specific
and threat-independent design provides the most robust structural system that is
able to survive both the anticipated and the unimaginable catastrophic events. As
with the performance-based design of structures to resist strong ground motions,
the design of structures to resist blast loads considers successive levels of damage and deformation in response to increasingly severe events. Also similar to
seismic design, the analytical methods that are required to demonstrate the performance of structural components and structural systems in response to extreme
blast loading are necessarily dynamic and inelastic.
9.1.1 Seismic versus Blast
Performance-based methods are used to design structures that must resist strong
ground motions and structures that must resist blast loading. In both cases, different extents of acceptable damage are specified for events of different intensity. This approach acknowledges the financial burden associated with comprehensive protection in response to the most extreme loading conditions and rationally accepts successively greater risk for successively more severe events.
This risk-based approach starts with a risk assessment that considers the function, criticality, occupancy, site conditions, and design features. However, unlike
natural hazards, man-made events are entirely unpredicable. As a result, risk
associated with blast loading can at best be quantified in a relative sense, in order to compare different levels of protection or different protective measures.
241
242
STRUCTURAL SYSTEMS DESIGN
Since there are not enough data to justify the absolute quantification of risk, the
choice of design basis threat is inherently subjective and is often based on past
experience.
The analytical methods that are used to evaluate the performance of structures in response to these forms of extreme loading involve the solution of the
equations of motion. Since ground motions are typically assumed to be applied
uniformly over the foundations of most structures, the spectral approach allows
the designer to envelop the expected response without having to consider a specific input motion time history. However, blast loads and blast effects are typically localized to the immediate vicinity of the detonation, and these enveloping
methods are not applicable. Localized blast loads will deform the exterior bays
of the structure to a much greater extent than the interior bays, and, in the time
frame of the applied loading, the story shears may not necessarily be distributed
through the diaphragms in proportion to the framing stiffness. Typically, the ductility demands at the blast-loaded face of the structure far exceed the demands
elsewhere throughout the remainder of the structure.
Some types of construction are inherently more ductile than others, and these
are better suited for seismic and blast-resistant design. While cast-in-place concrete can be detailed to provide sufficient ductility, post-tensioned two-way flatslab structures are problematic. Punching shear failures at the slab joint must be
avoided under lateral sway deformations, and these systems possess limited lateral deformation capability. Although a two-way flat-slab system might perform
satisfactorily in response to strong ground motion when it is used in combination with a stiffer shear wall system that controls the ultimate lateral sway of
the structure and prevents destabilizing P-Delta effects, the punching shear failure vulnerability is a function of the slab performance in response to blast infill
pressures and is not improved with the addition of shear walls.
The fundamental difference between seismic excitations and blast-resistant
design is the distribution of ductility demand throughout the structure. Strong
ground motions engage the entire lateral resisting system, while resistance
to blast loading is more concentrated in the structure in the vicinity of the
explosion. Moment frames may provide high ductility and energy absorption
when the plastic hinging of beams is spread throughout the structure. In
order for inelastic deformation to be distributed throughout the structure, the
beam-column connections must develop the flexural hinge in the beam. The
elements must also be proportioned to prevent plastic story mechanisms that
might limit the inelastic energy absorption to isolated regions of the frame. This
is achieved through the “strong-column, weak-beam” (SCWB) approaches for
both seismic and blast design. Since blast loading is more localized than seismic
excitations, the resulting plastic hinging is more localized, with more severe
rotations than those that result from strong ground motions. Structural systems
that effectively resist the lateral forces resulting from strong ground motions,
such as concentrically braced frames (CBF) and eccentrically braced frames
(EBF), must be considered on a case-by-case basis for use in blast-resistant
design.
GENERAL DISCUSSION
243
However, the manner in which seismic and blast forces are distributed
throughout the building is strongly affected by the size, shape, and location of
structural elements. Although seismic forces are proportional to the mass of the
building, inertial resistance reduces the response to blast loading. Structures that
are designed to resist both blast forces and strong ground motion benefit from
low height-to-base ratios, balanced resistance, symmetrical plans, uniform sections and elevations, the placement of shear walls and lateral bracing to maximize
torsional resistance, short spans, direct load paths, and uniform floor heights.
Although the benefits of employing these design elements may be similar, the
reasons for incorporating these design features may differ. For example, seismic
excitations may induce torsional response modes in structures with reentrant corners, whereas these conditions provide pockets where blast pressures may reflect
off adjacent walls and amplify the blast effects. Similarly, first-floor arcades that
produce overhangs or reentrant corners create localized concentrations of blast
pressure and expose areas of the floor slab to uplift blast pressures. Adjacent
structures may impact one another as they respond to base motions, while adjacent structures may be subjected to multiple reflections of blast waves. The
geology of the site influences the characteristics of the ground motion that load
the structure, and the geology influences the size of the blast crater and the reflectivity of the blast waves off the ground surface.
The ductility demands for both seismic and blast-resistant design require attention to details. Confinement of the concrete core prevents premature buckling
of the rebar and maintains load-carrying capacity, despite the significant cracking associated with large inelastic deformation. Closely spaced ties and spiral
reinforcement are particularly effective in increasing the ductility of a concrete
compression element. Except for near-contact detonations, carbon fiber wraps
and steel jacket retrofits provide comparable confinement to existing structures;
however, carbon fiber materials may be damaged when the standoff distances
are too short. Steel column splices must be located away from regions of plastic hinging or must be detailed to develop the full moment capacity of the section. Closely spaced stiffeners or concrete encasement will prevent local flange
buckling of steel sections. Reinforced concrete beam sections require equal resistance to positive and negative bending moments. In addition to the effects of
load reversals and rebound, doubly reinforced sections possess greater ductility
than singly reinforced counterparts. Steel beams may be constructed composite
with the concrete deck in order to increase the ultimate capacity of the section;
however, this increase is not equally effective for both positive and negative moments. While the composite slab may brace the top flange of the steel section, the
bottom flange is vulnerable to buckling. Tube sections and concrete encasement
are particularly effective in preventing flange buckling under load reversals.
9.1.2 Analytical Methods
A wide variety of analytical approaches may be used to determine the performance of structures in response to explosive loading. The various methods
244
STRUCTURAL SYSTEMS DESIGN
include empirical explosive test data that can be expressed as prescriptive
requirements or pressure-impulse charts, single-element response analysis,
structural system response analysis, and detailed finite element analyses. These
dynamic response methods all characterize the behavior of the structural
system in response to a short-duration but high-intensity dynamic loading.
The dynamic analysis methods must be able to capture the frequency of the
loading function and the dominant frequencies of the structural system, and
represent the governing failure mechanisms so as not to overlook critical failure
mechanisms that may precipitate collapse or hazardous debris. The designer
must therefore anticipate the likely failure mechanism before selecting the
appropriate analytical approach.
9.2 MODELING
Long before complex finite element analysis methods were readily available, the
design of structures to resist blast loading relied extensively on single-degree-offreedom (SDOF) methods. These methods were generally conservative as long as
the designer anticipated the system response and made sure the analytical model
captured all failure mechanisms and response characteristics. Advanced finite element methods may be used where structural complexity or spatial distribution of
loading requires more detailed analyses. These methods can represent complex
geometries, connectivity, and material properties. While simplified models are
prone to problems associated with too little information, complex models need
to be tailored to available resources and time constraints. The optimal model
therefore contains enough information to capture the response of interest without becoming so excessively complex that it cannot solve the equations of motion within a reasonable solution time. Accepted practice requires that enough
resolution be included in the model to accurately capture response modes.
When fast-running simplified methods are deemed inadequate, detailed
analytical models can often be constructed to calculate structural performance in response to blast and impact. The current state of the art relies on
explicit dynamic inelastic finite element software such as NLFlex, LS-DYNA,
AUTODYN, Abaqus, Pronto, and the like. When modeling a significant section
of a building, such as a couple of bays, the required level of resolution or
element size to calculate an accurate response on current-generation computer
hardware limits the extent of the structure that can be analyzed. For these 3D
models, the required computer RAM increases by the cube of the resolution,
and the computational time increases by the 4th power. As a result, the analysis
of larger models will become feasible as computer hardware advances.
At this resolution level, each rebar is typically represented explicitly with
beam or bar elements, and structural steel is typically represented explicitly with
shell or plate elements. Concrete, soil, and rock are represented with continuum
elements. In view of this background, macroscopic phenomenological constitutive models are the state of the art for common construction materials such
MODELING
245
as steel and concrete. Research efforts on more fundamental meso- and micromechanical approaches fill the literature; however, these are currently too computationally demanding for production use.
The goal of a phenomenological model is to accurately capture the important
characteristics of material response in the expected loading regime, while obeying the laws of thermodynamics. The theory must be implemented in a robust
algorithm that is efficient in terms of both RAM and CPU usage. Important material characteristics for concrete include strength dependence on pressure, energy absorption or hysteresis due to pore compaction, increase of resistance with
loading rate, and softening due to fracture or crushing. Concrete also transitions
from a brittle material that fractures at low pressures to a more ductile material
that flows at higher pressures. Important considerations for steel include hardening after yield, strength dependence on loading rate, and fracture dependence on
confining pressure. A realistic representation of fracture or material softening is
very important in blast applications. Once softening initiates, material damage
localizes into discrete cracks or narrow bands. Maintaining accurate simulations
during localization requires some sort of regularization. At currently achievable
resolution levels, this usually requires a fracture energy-based approach whereby
the softening stress-strain response is a function of element dimension.
Theories of plasticity, viscoplasticity, and/or continuum damage mechanics
are a few of the popular theoretically sound frameworks for phenomenological
models. A well-formulated model will use parameter identification from a small
number of lab tests to reasonably characterize material response over a very wide
range of loading conditions.
9.2.1 Systems
Several of the most common structural systems will be discussed, including the
significant features that must be represented in the analytical models. While this
list represents a sampling of structural systems that may be analyzed, it is not intended to limit the types of structural systems that may be used for blast-resistant
design.
Concrete and Steel Moment Frames Moment frame buildings may be analyzed one component at a time, modeling each component as an SDOF system
with dynamic blast loads either applied tributary to the loaded surface area or collected through the reaction forces from secondary members as they frame into
the primary structure. The appropriate performance limits for the calculated inelastic deformations are typically conservatively specified, particularly for axial
members, where P-Delta effects may initiate structural instability. Furthermore,
simplified SDOF methods are incapable of capturing localized steel flange deformations, steel splice performance, concrete panel zone deformations, and other
localized response characteristics. These SDOF approaches are only appropriate for standoff detonations that apply fairly uniform loads to the member. The
overall frame response may similarly be determined using an SDOF model of
246
STRUCTURAL SYSTEMS DESIGN
the building’s lateral resisting system. These blast-induced base shears recognize
the flexibility of the frame and a characteristic extent of frame ductility; however,
since blast damage is typically localized in proximity to the point of detonation,
these overall ductility levels are approximate and represent overall energy dissipation of the framing system. Fundamental to the modeling and analysis of
the frame using these methods is the appropriate detailing of the connections so
the structural members will be capable of developing the calculated extent of
rotation and lateral drift.
Steel-Braced Frames and Concrete Frames with Concrete Shear Walls Similar to the response of concrete and steel moment frames, the response of steelbraced frames or concrete frames with concrete shear walls may be analyzed one
component at a time, to determine local response in proximity to the detonation;
however, the global response of the structure will be characterized by a stiffer
lateral resisting system with a higher fundamental frequency of response. Not
only will the period of the structural system be shorter than that of a comparable
moment frame structure, but the amount of energy dissipation will be less. This
type of structural system will typically develop larger blast-induced base shears,
and the floor diaphragms must be capable of distributing the collected lateral
loads to the various lateral resisting components.
Precast Tilt-Up with Concrete Shear Walls Precast tilt-up load-bearing walls
may be modeled as single-degree-of-freedom systems, provided the detonation
is sufficiently far from the building to produce a uniform loading over the floor
height. The flexural deformations of these wall systems must be limited in order
to prevent P-Delta effects that may precipitate instability, and the diaphragms
must be sufficiently tied to the wall in order to transfer the large lateral loads and
rebound forces. Subsequent diaphragm models are required in order to demonstrate their ability to distribute loads to the concrete shear walls.
Reinforced Masonry Walls with Rigid Diaphragms Reinforced masonry
walls may be analyzed as single-degree-of-freedom systems; however, pressureimpulse diagrams that represent empirical data and the results of analytical studies may be used to evaluate their performance in response to blast loading. As
with precast tilt-up construction, analyses must be performed in order to demonstrate that the diaphragms are capable of distributing the lateral loads to the shear
walls and are able to restrain the large rebound forces.
9.2.2 Materials
Structural materials are affected by extremely high loading rates, high confining pressures, and large inelastic deformations in response to blast loading.
Simplified elastic-plastic resistance functions are often used to represent
ductile materials with dynamic enhancement factors that account for the strain
MODELING
247
rate loading. Explicit dynamic inelastic finite element analyses require more
sophisticated material models to represent the behavior of structural material.
Materials such as concrete and steel are typically stronger than the minimum
values specified on the construction documents. Although ASTM specifications
and codes define the minimum properties for materials, the actual materials being
installed typically have strengths that exceed these minimum requirements. For
the analysis and design of reinforced concrete and structural steel elements for
flexure, it is generally acceptable to increase the steel material’s minimum yield
strength by 10%. Ignoring the material’s average strength and dynamic strength
increase generally results in increased factors of safety for bending elements;
however, underestimating the effects of material overstrength is nonconservative
for shear and connection design.
Structural elements also exhibit higher strengths when subjected to dynamic
blast loads than they exhibit in response to static loads. Material testing demonstrates an increase in strength in response to high-strain-rate dynamic loadings.
As with the use of average strength factors, underestimating the high-strainrate effects produces greater factors of safety for flexural element design but
is nonconservative for shear and connection design. However, the factored yield
strength that will be used for determining the flexural resistance function should
not exceed the ultimate dynamic strength of the material.
Threat-Dependent Effects Explosive threat can be grouped into three general
categories based on the explosive weight and standoff, contact/near contact, and
near field and far field. Contact and near contact place the explosive threat in
close proximity to the structure, generally at scaled ranges less than 0.5, scaled
standoff range (Z = R/W 1/3 ), where R is standoff, and W is explosive weight. At
this standoff distance, the explosive effects are highly localized, and the air-blast
loads have extremely high-pressure short-duration pulses. At these scaled ranges,
use of simplified tools is questionable. Scaled ranges between 0.5 and 2 fall into
the near-field range. If charge shape is not an issue and the structure surface is
flat and regular, simplified air-blast loading functions are applicable. Far-field
threats generally exist at scaled ranges greater than 2.0. Pressure loading is more
uniform across the structure with lower peak pressures and long duration pulses.
Air-blast clearing can be a significant issue for far-field threats.
Near Range versus Far Range The amplitude of peak pressures for explosive
detonations that are in close proximity to a structure or structural element is extremely high. At these short distances the blast loading is taking place within
the explosive by-products, and the reflection factors may be in excess of 10. Although the duration of the applied loads is short, particularly where complete
venting of the explosion occurs, the corresponding impulses are also extremely
high. Fragments associated with the high-pressure range usually consist of highvelocity primary fragments from the explosive casing breakup or the acceleration
of objects positioned close to the explosion. Simplified analytical tools that are
based on the modified Friedlander equations (Hyde 1992, Kingery and Bulmash
248
STRUCTURAL SYSTEMS DESIGN
1984) become less reliable for scaled ranges that are less than 0.37 to 0.45. Advanced analysis and testing show that when charge shapes other than spheres or
hemispheres are considered, simplified methods lose accuracy at scaled ranges
as high as 2.0. Advanced analytical methods based on computational fluid dynamics (CFD) are required to calculate accurate loading for near contact or for
complicated surfaces.
For near contact, the durations of the applied loads are much shorter than the
time it takes the structural elements to reach maximum deflections; they can often be analyzed for the impulse rather than the peak pressure associated with
longer-duration blast pressures. Also associated with the “close-in” effects of a
high-pressure design range is the possible occurrence of spalling and scabbing
of concrete elements and post-failure fragments. Spalling and scabbing result
from the disengagement of the concrete cover over the reinforcement on the
back side of a concrete structure (the side away from the explosion). Spalling
is caused by the high-intensity-blast shock wave reflecting off the back surface as a tension wave, and scabbing is associated with the disengagement of
concrete as it undergoes large displacements. Secondary fragments that result
from spalling and scabbing of structural materials can be a hazard to people and
equipment.
Structures that are vulnerable to close-in detonations are vulnerable to breach
and direct shear. Protective and containment structures adjacent to occupied areas should be constructed with double walls separated by a void of sufficient size
to allow for the deflection of the outer sacrificial wall. The relative strength of the
two walls and the required space between them may be determined on a case-bycase basis. Although simplified approximations may provide design guidance,
explicit dynamic nonlinear finite element analyses are required to accurately determine the potential for breach and spall.
Where scaled ranges are large, the duration of the blast loading may exceed
the response time of the structure. Dynamic or quasi-static response characteristics may be observed for this type of loading. Simplified single-degree-offreedom methods may accurately determine flexure and flexural shear failures
that may result from far-range loading.
9.2.3 Members
This section will discuss the modeling of structural components, such as
columns, beams, slabs, and walls, to determine their behavior in response to blast
loading. Included in this discussion will be the evaluation of deformation and reaction forces associated with flexure, shear, and axial effects. Also discussed will
be issues such as breach, spall, and fracture.
In a numerical model, the more accurate the representation of a member, the
better the results will be. Modeling of structural components, such as columns,
beams, slabs, and walls, can be done on the component level, as an assembly or as
a complete structure. When they are modeled on the component level, care must
be taken to prevent the results from being adversely impacted by the assumed
MODELING
249
boundary conditions. Generally boundary conditions will be either free, fixed,
pinned, or a combination. It is good practice to vary the boundary conditions to
see the impact each choice has on the response of the component. Fixing the
axial response at both ends of a member will result in a stiffer section than a
member with a fixed and a free end. Application of boundary conditions directly
to the ends of a member can result in additional confinement to pressure-sensitive
materials, such as concrete. This additional confinement can prevent shear failure
and force a member into a flexural response. Single-component models allow
the analyst the opportunity to perform grid refinement studies to determine the
impact of resolution on the accuracy of the solution. While single-component
models are the simplest to construct and the fastest to run, the results may not
provide the required accuracy.
The next step up from a single-component model is an assembly model. Examples would be a column with the addition of bracing beams, or a wall with
beams and columns surrounding. The increased model size and run time for
assembly models can improve accuracy due to more realistic boundary representation. The boundary conditions are further removed from the component of
interest, resulting in a better representation of the physical system. The increased
complexity of assembly models necessitates the introduction of connection details. This is especially true in modeling of steel-frame structures.
Full-structure models provide the greatest accuracy in modeling of a structure
if the level of refinement can be maintained. The trade-offs among model size,
run time, and level of refinement usually result in models of a full structure having a lower level of refinement. The reduction in refinement is traded off versus
the more accurate boundary conditions and the addition of the overall response
and flexibility of the full structure.
Advanced finite element methods using accurate material models are capable
of capturing the response of members through failure. Level of damage can be
obtained directly from the analysis results by evaluating deformations, velocities,
strains levels, and material damage metrics. Simplified analysis approaches rely
on guidelines based on comparison with testing or advanced analysis to relate
displacements or rotations with level of damage. These guidelines, which can
be found in military manuals and design references, are generally conservative
(Unified Facilities Criteria 2002, Unified Facilities Criteria 2008b). This will
better enable a structure to survive the applied load. The available test databases
are not extensive enough to cover all possible design configurations; as a result,
the accuracy of predicted damage can be affected.
Shear and Flexural Reaction Forces For most structural members, their response when subjected to a blast loading is a combination of shear and flexure.
How the member responds is a function of the structural component’s properties and the intensity of the loading. Long, slender members are more likely to
respond in flexure than short, thick sections. The intensity of the load also impacts the type of response. A more impulsive loading is more likely to produce
a shear response, as compared to a longer-duration load. As a result, a column
250
STRUCTURAL SYSTEMS DESIGN
that responds primarily in bending due to an explosive threat at a large standoff
distance may respond in shear due to a near-contact blast.
Reaction forces collected at the boundaries of component models can be extreme. Reactions from shear deformation modes are usually of short duration
with a high peak force. Reactions from primarily bending modes have a longer
duration on the order of the period of the member. Whether a section deforms in
shear or flexure depends on the properties of the section and the applied load.
Inclusion of supporting members or modeling of a region extending beyond
the component of interest will result in more accurate representation of reaction
forces. When the compliance of bracing members is included in the modeling,
the reaction at the column-beam intersection will be reduced. This allows for a
more efficient and cost-effective design of a connection.
Axial Effects Accounting for large deformation effects such as arching, stress
stiffening, tension membrane, and P-Delta buckling is critical to the accurate
evaluation of structures subjected to blast. Simplified modeling approaches, such
as SDOF, do not account for large deformation effects in their resistance functions. As a result, the prediction of response is usually based on rotations or
displacements, and the level of damage is evaluated based on guidance from calibration of the models to test results or an advanced analysis database. Extrapolation outside of the available test database can result in inaccurate assessments of
structural damage. Advanced finite element methods are formulated to account
for geometric nonlinear and large deformation effects and are better suited for
evaluating response over the full range of damage through collapse.
The additional confinement provided by axial preload can influence the
response of concrete structural components such as columns or load-bearing
walls subjected to blast loading. Strength dependence on confining pressure
can increase the capacity of concrete materials subjected to blast loading. In
general a vertical structural component loaded with a lateral blast pressure will
initially experience a compression membrane, followed by a tension membrane
as the deformations increase. During compression membrane behavior, the
structural resistance to lateral deformation may increase. However, as the section
deforms due to blast loading and the lateral deformation grows, the ability to
resist deformation transitions from compression arching to tension resistance.
Post-blast damage stability is dependent on the damaged section’s ability to
continue to carry axial load. Geometric P-Delta effects can result in collapse
under axial loads.
The response of steel columns subjected to blast is less sensitive to initial axial
load. The addition of an axial preload does have an impact on the post-damage
failure of the section or collapse due to P-Delta effects. A common method for
evaluating the response of axially loaded members is to break the response up
into load phases. Initially the member is preloaded with the axial force. The top
axial degree of freedom is fixed, locking in the preload. The member is then
subjected to the lateral blast load, and the damage to the column due to blast is
evaluated. The axial degree of freedom is then released to allow either the axial
MODELING
251
force on the column to stabilize or the damaged column to collapse under the
axial load.
Fracture Capturing fracture of a member requires the use of high-fidelity
modeling. Simplified methods such as SDOF models can be calibrated to
testing or high-fidelity modeling results to give indications of when fracture
occurs. Displacement or rotation limits can be used to predict failure, but these
are usually conservative. Advanced finite element methods, using constitutive
models that capture the response of materials undergoing large strain and
softening, are required to investigate fracture of members. The evaluation of
breach and spall is impacted by the model’s ability to calculate fracture in the
material. If fracture is not accurately modeled, the ability to model breach and
spall will be compromised.
9.2.4 Connections
This section will discuss the modeling of structural connections in order to determine their behavior and performance in response to blast loading. Included in
this section will be the effects of shear and flexure associated with small deformations and the axial forces generated by large deformations.
Modeling of connections is critical for accurately predicting the response of
structures subjected to blast and the potential for progressive collapse following
a blast. Evaluation of connections is best accomplished using advanced modeling
approaches. The complicated nature of connections requires a refined treatment
of the details of their construction. For concrete structures, the additional confinement provided by the reinforcing steel in the joint can increase capacity and
ductility. Models of a concrete joint need to include the detailed geometry and
accurate material models that can account for the strength increase resulting from
this confinement. Steel connections can be complicated and require the modeling
of contact surfaces between members and connection plates and angles. Failure
of welds or bolts can dominate the response of joints, and detailed models are
required to address possible failure modes including fracture of welds or surrounding steel, bolt failure, and bearing failure of bolt holes.
Large Deformations The response of a connection changes as it deforms. Rotations can become large, and the resulting forces at the joint change. Initially at
small deformations the forces in a joint can be dominated by shear and compression. As the joint deforms more, the method of force transfer through the joint
can change to tension. The forces in the connection are impacted by a number
of effects, such as prying, twisting, fracture, failure in bearing, and local buckling. These complicated geometric and material effects are difficult to capture
accurately. Details of connection construction are difficult for simplified tools to
model accurately.
Shear versus Flexural Effects During blast loading, a connection can experience direct blast loads and loads transferred from neighboring members. As
252
STRUCTURAL SYSTEMS DESIGN
adjacent members fail, additional forces can be transferred through connections
from surrounding members. These forces can be transferred by shear and flexure. Bending of neighboring members can result in lateral forces due to tension
or compression of the member. Prying at the joint can also induce lateral loads.
9.3 ANALYTICAL APPROACHES
The analytical method that may be used to evaluate a structure’s response to
short-duration but high-intensity dynamic loading depends on the characteristics
of the loading and the anticipated failure mechanism. These approaches range
from simplified table lookups to multi-degree-of freedom finite element calculations, and each approach has inherent constraints or limiting assumptions that
must be appropriate for a specific application. For example, blast-resistant glass
thickness may be selected using ASTM F2248, Standard Practice for Specifying
an Equivalent 3-Second Duration Design Loading for Blast Resistant Glazing
Fabricated with Laminated Glass (ASTM 2003), and ASTM E1300, Standard
Practice for Determining Load Resistance of Glass in Buildings (ASTM 2007).
Empirical explosive test data can often be expressed as prescriptive requirements
or pressure-impulse charts. Flexural response behavior of structural components
can often be determined using single-degree-of freedom methods. The Component Explosive Damage Assessment Workbook (CEDAW) (Oswald 2005) and
Single-Degree-of-Freedom Blast Effects Design Spreadsheets (SBEDS) (Protective Design Center 2006) were developed by the U.S. Army Corps of Engineers
(ACE) for use in blast effects assessments and utilize the pressure-impulse (P-I)
and single-degree-of-freedom (SDOF) response methodologies. Detailed finite
element analyses are computationally intensive and require complex modeling
of the structure and material properties. The choice of the most efficient analytical approach requires a full understanding of the benefits and limitations.
In all cases, the analytical methods must be numerically robust in order to
characterize the transient blast loading pulse and the dominant frequencies of
response associated with the postulated failure mechanisms. Care must be taken
so as not to neglect the critical failure mechanisms that may precipitate collapse
or hazardous debris. Since it is important that the selected analytical approach
be capable of representing the structure’s likely failure mechanism, the designer
must be able to anticipate the response when using simplified analytical methods.
This is not the case for a finite element approach in which the model should
capture all the failure modes.
9.3.1 P-I Diagrams
Pressure-impulse (P-I) diagrams summarize the performance characteristics of
a type of structural or nonstructural element in response to a range of explosive
loading. The explosive loading parameters are expressed in terms of peak pressure and impulse, and the corresponding P-I curves indicate the threshold of the
ANALYTICAL APPROACHES
Figure 9.1
253
Pressure-Impulse (P-I) Diagram for Different Levels of Damage
different levels of protection or hazard. See Figure 9.1. Typically, one end of the
concave P-I curves transitions to a pressure asymptote, while the other end of
the curve transitions to an impulse asymptote. These two asymptotes represent
the minimum impulse for all greater peak pressures or minimum peak pressure
for all greater impulses that correspond to a specified performance threshold. All
combinations of peak pressure and impulse that lie below or to the left of a P-I
curve correspond to a better performance condition, while all combinations of
peak pressure and impulse that lie above or to the right of a P-I curve correspond
to a worse performance condition. The P-I curves either may be derived from an
analytical database of performance or may be fit from available test data. Often,
the parameters are presented in nondimensional form where the peak pressure
and impulse axes are normalized so that a given set of curves and plots may be
used to represent the performance of a greater range of element responses to a
wide variety of explosive threat conditions.
9.3.2 Single-Element Analyses
Individual elements may be analyzed independently by means of either singledegree-of-freedom (SDOF) or multi-degree-of-freedom (MDOF) inelastic dynamic methods. These analytical methods require the use of explicit time-step
numerical integration algorithms, ranging from constant velocity to linear acceleration techniques, along with the appropriate constitutive relations that represent both the elastic and inelastic behavior of the element. SDOF models are
254
STRUCTURAL SYSTEMS DESIGN
inherently simple, and the accuracy of the response calculation depends on the
approximations that are used to characterize the dynamic response of the element. MDOF models may be significantly more complex and may be developed
using finite element methods (FEM) to derive the mass and stiffness matrices.
While it is permissible to consider damping effects, these effects are generally
small in comparison to the large inelastic response that may result from blast
loading and are often neglected.
SDOF SDOF methods are adequately described in a variety of structural
dynamics textbooks (Biggs 1964) and U.S. Army Technical Manuals (Unified
Facilities Criteria Progam 2008b). This approach is based on solving the equation
of motion for a spring-mass system with a nonlinear resistance function using numerical integration. The approximate design methods are based on a RayleighRitz approach for which representative shape functions for the flexural mode are
assumed and used to calculate the effective mass, effective stiffness, and effective
loading function. When the loading pulse is very short relative to the period of the
dynamic system, the initial velocity may be approximated as impulse divided by
the effective mass, and the kinetic energy of the system can be equated to the area
under the force-displacement resistance function. This provides a closed formed
relationship between the required ultimate resistance (Rm) , the applied impulse
(I), the frequency of the dynamic system (ω) and the allowable ductility (µ).
Rm = I ω/(2µ − 1)0.5
The accuracy of these approximate approaches depends on selecting the appropriate SDOF system to represent the governing failure mechanism of the element. While this approach is most commonly used for the design and analysis of
elements that are subjected to blast loading, it requires considerable experience
to make sure a critical failure mechanism is not overlooked. Typically, flexural
modes of response are represented by the SDOF models; however, direct shear
and instability may prove to be more critical response characteristics for shortstandoff detonations and for axially loaded columns. The dynamic loads that may
be applied to SDOF analyses of individual structural elements may be based on
tributary areas or may be derived from the dynamic reactions of the secondary
elements that frame into a primary element. The application of dynamic reactions from subsidiary elements accounts for the frequency of response and limiting capacity of the subsidiary elements, but must also include the appropriate
amount of mass associated with the subsidiary element response. The dynamic
reaction forces from a series of subsidiary element analyses may be converted
into an equivalent dynamic load by equating the work performed by the individual reactions through a representative deformation to the work performed by the
equivalent SDOF response motion. Therefore, if SDOF analyses are performed
for individual secondary beams or purlins, the calculated reaction forces at every time-step may be multiplied by the shape function amplitude of the primary
beam or girder at the location of their connections. If this quantity of work is
ANALYTICAL APPROACHES
255
divided by the integral of the shape function over the span, the resulting quantity
is the effective uniform load at the corresponding time-step that is applied over
the entire span.
MDOF MDOF analyses may be performed using a variety of dynamic response analysis software programs; however, all analyses must be performed
at a sufficiently short time-step to guarantee numerical stability, to resolve the
frequency content of the model, and to accurately represent all potential modes
of failure. Nonlinear dynamic finite element analyses provide the most comprehensive means of determining the self-consistent stiffness and inertial properties,
the limiting plastic limits, and the governing failure mechanisms. Furthermore,
nonlinear geometric algorithms provide the means to include the destabilizing
P-Delta effects due to the coupling of the large lateral transient deformations
and the static gravity loads. MDOF models accurately represent the spatial distribution of dynamic pressures and dynamic reactions from subsidiary element
analyses without having to generate an equivalent dynamic load.
9.3.3 Structural Systems Response
SDOF responses and MDOF models of individual elements do not account for
the interaction between interconnected elements, the phasing of their responses,
or the flexibility of the actual boundary conditions. These simplified approaches
rely on engineering judgment for the selection of approximate boundary conditions, such as soft springs and additional mass. As a result, the idealizations
represented by SDOF and MDOF analyses of individual structural elements do
not accurately account for the dissipation of energy as the entire structure is
deformed and suffers damage. Structural system response calculations include
the relative flexibility and strength of the interconnected structural elements and
provide a more accurate distribution of blast loading. Most importantly, however, structural system analyses consider the phasing of the responses between
the different structural elements. Often, the region of greatest interest is modeled in greater detail, while the rest of the structure is more coarsely represented.
This allows the analyst to dedicate the greatest computational resources to the
response of critical details. However, as with all dynamic response analyses, the
transition from coarse representations to more detailed models must be gradual
so as not to generate spurious results.
9.3.4 Explicit Dynamic Finite Element Analyses
Explicit dynamic finite element methods require sufficient modeling resolution
and short enough time-step to capture the high-frequency characteristics of the
shock wave loading and structural response. Explicit formulations of the equations of motion express the displacement at a given time-step tj + 1 (where the
indices j represent the increments or steps in time for the duration of the response
analysis) in terms of displacements, velocities, or accelerations at previous
256
STRUCTURAL SYSTEMS DESIGN
time-steps. This approach captures the high-frequency effects of the shock loading and easily allows for both material and geometric nonlinear effects; however,
all explicit formulations have a critical time-step, above which the solution becomes numerically unstable. The critical time-step is related to the wave speed of
the material and the least dimension of the finite element model. The equation of
motion is therefore solved at each node, using the current geometry and material
properties at that location and point in time.
An implicit formulation of the equations of motion expresses displacement
of each node at a given time-step tj + 1 in terms of all the displacements, velocities, or accelerations at that time-step tj + 1 . The equations of motion for the
entire system must be solved simultaneously; this is usually done using matrix
methods and is a very efficient way of calculating the response of linear elastic
systems. For nonlinear dynamic systems, however, an explicit formula (predictor) is usually used to estimate the response at the end of each time-step, and this
is followed by one or more corrections to improve the results. Such an approach
is sometimes very inefficient. Furthermore, even for elastic analysis, if the shock
or high-frequency response of the system is needed, the time-step must be made
sufficiently small to capture such effects while avoiding undue numerical damping. Implicit schemes are least efficient in such cases.
Given the complexity of using explicit dynamic finite element methods, great
care needs to be exercised to achieve sufficient accuracy. Verification and validation of both the software and the analyst are critical to the successful completion of high-fidelity first-principle calculations. Comparison of analysis methods
versus physical test data is one method for validating computational tools and
approaches. Benchmarking versus known analytical solutions or against other
software packages provides further confidence in software and methods. Unless
established methods are used based on validation exercises, finite element methods need to be checked with grid refinement calculations to determine whether
the model has sufficient resolution. Fine resolution and detailed models are no
guarantee of accurate results. Accurate representation of material response, sufficient physical extent of the computational model, and proper boundary conditions are critical to calculating accurate results.
9.4 PROGRESSIVE COLLAPSE
The protective design of structures is often referred to as threat-dependent or
threat-independent. In one case, a structure is designed to provide a specified
level of performance in response to a specific design basis threat, while in the
other, the structure is designed to be insensitive to local damage resulting from
unforeseen extraordinary events. In many cases, the uncertainty of the maximum credible threat or the inability to accurately characterize the performance
in response to explosive loading requires both approaches to be applied to the
design. While the threat-dependent approach is related to the design basis threats,
the threat-independent methods evaluate the potential for an initiating event to
PROGRESSIVE COLLAPSE
257
precipitate a broader extent of collapse. The prevention of progressive collapse
may therefore be achieved through the hardening of key structural elements in
order to preclude the initiating event, through the provision of ductile redundant
load paths, or through compartmentalization that permits local failure to occur
without endangering adjoining structural elements. In all cases, consideration
must be given to both global ductility demands that involve all elements of the
framing system and local ductility in the vicinity of the initiating damage.
The insensitivity to local failure, often referred to as robustness, is a property
of the structure, and is independent of possible causes of initial local failure.
Where specific threats are not prescribed, or for structures that are deemed low
risk, an assumed extent of initial local failure and assumed extent of damage
to the remaining structure may be prescribed. For these structures, element-toelement connectivity, minimum tie forces, the elimination of nonductile failure
mechanisms, and enhanced stability requirements may improve the structure’s
robustness. If the structural members and connections can develop these minimum tie forces, which vary with construction type and location in the structure,
the structure will be held together in the event local damage disrupts the primary
load paths. The purpose of the horizontal and vertical ties is to enhance continuity and ductility, and to develop alternate load paths in the structure. Internal
and peripheral horizontal ties are typically provided, along with ties to external
columns and walls. The extent of tie force requirements may either be prescribed
or calculated, starting with an assumed damage pattern. The alternate load path
requirements, also termed “bridging requirements,” may be determined using dynamic nonlinear analytical methods and known material damage limits, or may
be approximated using equivalent static elastic analysis methods with nominal
material damage criteria. The level of conservatism may be varied depending on
the level of approximation or the risk assessment for the structure.
Where the characteristics of the structure make bridging impractical, local
hardening of key elements or the introduction of structural fuses may be introduced into the structure to either prevent the initial damage or define a limited
zone of collapse. For these types of structures, the level of specific local resistance for key elements or the zones of limited collapse must be defined.
Where a threat and risk assessment quantifies the hazards to which a structure may be subjected, specific initiating events may be prescribed, and the performance of the structure may be calculated. Nonstructural protective measures
may be developed to deny access to key structural members, or details and sacrificial elements may be introduced to isolate the critical components. For this
performance-based approach, the structural system redundancy and element capacities would be designed to resist these initiating events or resist the progression of damage or collapse. The increased resistance is therefore based on a prescribed set of actions and a permissible extent of resulting damage. The intensity
of the initiating actions and the permissible response limits may be specified in
the security performance criteria for the structure. The analytical methods that
demonstrate the behavior of the structure must accurately represent the dynamic
nature of the event and the damage-state behavior of the structural materials.
258
STRUCTURAL SYSTEMS DESIGN
Regardless of whether local hardening, bridging over a damaged structure, or
compartmentalization is used, a post-event stability analysis of the structural system must be performed.
For blast-resistant design, a threat must be identified in order for the initial
state of damage and the residual capacity of the structure to be determined. This
approach is similar to that of performance-based seismic design codes for which
the engineer selects the hazard (event) and the desired performance, such as low
levels of damage (immediate occupancy), in response to a small earthquake and
life-safety protection (damage short of complete collapse) in response to an extreme event. Either the structural elements must be designed to resist the specified threat and maintain load-carrying capacity, or a finite extent of localized
damage may be permitted, as long as the damaged structure is still capable of
preventing a progression of collapse. This corresponds to the “extreme event” in
response to which the structure must remain standing for life safety and evacuation of the occupants.
9.4.1 European Guidance
In the Euro-Norm EN 1991-1-7 (British Standards Institute 2006), the strategies
for preventing collapse consider two different approaches: strategies based on
identified accidental actions and strategies based on limiting the extent of localized failure. The strategies based on accidental actions consider designing the
structure to have sufficient minimum robustness, protective measures that prevent the action or reduce its effects, or designing the structure to resist the effects
of the action. The strategies based on limiting the extent of localized failure address the “threat-independent” prescriptive approaches for robustness, such as
enhanced redundancy through alternate load paths, key element designs to withstand notional accidental action, and prescriptive rules for integrity and ductility.
However, the criteria for protection are specific to the project and are determined through consultation with the client and the relevant authority. Since
strategies based on unidentified accidental actions cover a wide range of possible events, the means to prevent progression of collapse are typically based on
limiting the extent of localized failure. The magnitudes of accidental actions that
are identified, as in the case of internal explosions and impact, are specified in
the EuroNorms. In some cases where there is no risk to human life and where
economic, social, or environmental consequences are negligible, the complete
collapse of the structure caused by an extreme event may be acceptable.
9.4.2 U.S. Guidance
All current U.S. Guidance documents are based on the U.K. standards that were
based on the partial collapse of the Ronan Point apartment house. These standards extrapolated from the behavior of a panelized precast structure to develop
prescriptive and performance-based guidelines for other forms of construction.
Similar requirements form the basis for the most recent Building Code of the
PROGRESSIVE COLLAPSE
259
City of New York (New York City Department of Buildings 2008), the 2008 Supplement to the International Building Code (International Code Council 2008),
and the U.S. Department of Defense’s UFC 4-023-03 (Unified Facilities Criteria 2008a). Since these approaches all rely on catenary action, the structural
detailing must consider connections that are able to undergo significant inelastic
rotations while sustaining large tensile axial loads.
Since the alternate path method is inherently dynamic and inelastic, the
optimal analytical approach considers material and geometric nonlinearity and
inertial effects. While the analysis of dynamic systems may be performed using
implicit or explicit numerical integration algorithms, the introduction of material
and geometric nonlinearity complicates the analysis. Implicit schemes require
the reformulation of the stiffness matrix whenever properties are significantly
altered, and this can become computationally inefficient when significant extents
of inelasticity or catenary behavior are introduced. Explicit schemes require
small enough time-steps to ensure numerical stability, and this too may become
computationally inefficient. Furthermore, unless the beam-column joints and
connections are explicitly modeled, their behavior must be approximated with
equivalent nonlinear rotational and translational springs. The force displacement
relations for these equivalent nonlinear springs are best determined by performing comprehensive component analyses of specific joints, as was done for the
analysis of the Deutsche Bank Building for the World Trade Center Building
Performance Study (Federal Emergency Management Agency 2002).
The most recent version of UFC 4-023-03 (UFC, 2008a) proposes a simplified
procedure that is based on the approaches established for Seismic Rehabilitation
of Existing Buildings (American Society of Civil Engineers 2007). A ductile
material that develops significant deformation following yield, over which the
force is either constant or increasing, is said to exhibit deformation-controlled
behavior. Alternatively, a brittle material that experiences a steep reduction
in capacity following limited inelastic deformation exhibits force-controlled
behavior. The representations of structural behavior that are reported in ASCE
41-06 are considered to be reasonable for the evaluation of damage-state performance of buildings. Although simplifications to the dynamic inelastic response
analyses reduce the extent of computational resources needed to calculate the
performance of the structure following the removal of a key element, these
simplified methods require detailed guidance in order to capture all reasonable
failure mechanisms. UFC 4-023-03 offers two simplified static analysis procedures, linear static procedure (LSP) and nonlinear static procedure (NSP), as
alternatives to the nonlinear dynamic procedure (NDP) approach. These simplified approaches allow the use of commercially available structural analysis
software that is readily available in structural design offices. The simplifying approximations allow the analyst to apply load factors to static analyses in order to
account for the missing dynamic effects, and provide the means to interpret the
results of linear analyses to account for the material and geometric nonlinearities. Since the NDP is the most accurate representation of the structural response
to an initial state of damage, there is little ambiguity to the interpretation of the
260
STRUCTURAL SYSTEMS DESIGN
calculated results. Correspondingly, all approximate methods are necessarily
conservative, and a more representative analytical understanding will allow the
designer to achieve a specified level of robustness using a wider variety of structural options. The use of the simplified LSP is restricted to buildings with regular
framing plans, which are defined as irregularity limitations. Both approximate
methods were calibrated to nonlinear dynamic analysis to establish simplified
system-dependent factors that account for the anticipated dynamic response.
The LSP may be used wherever structural irregularities and demand capacity ratios conform to prescribed limitations. For the LSP, the structural model
is analyzed in response to both deformation and force-controlled actions. The
actions applied to the tributary areas above the removed key element are amplified to account for the inertial effects associated with the sudden removal of
the column, and a concurrent lateral load is applied individually to each face
of the structure to evaluate stability. The LSP may still be used if the structural
framing is irregular, provided the calculated demand/capacity ratios (DCR) are
less than or equal to 2.0; otherwise the NSP must be used. Once the acceptable use of the LSP is determined, the potential for disproportionate collapse
is calculated using component or element demand modifiers (m-factors) in an
approach that is similar to that of ASCE 41-06. M-factors are tabulated for different types of structural systems and details for different levels of damage-state
performance. These m-factors are used to determine both the Dynamic Increase
Factors and the maximum acceptable forces for deformation-controlled actions.
The expected material strength and the specified material strength respectively
define material properties for deformation-controlled and force-controlled analyses, and the expected component strength (QCE ) and lower bound estimate of the
component strength (QCL ) respectively define action capacities for deformationcontrolled and force-controlled analyses. QUD is defined as the deformationcontrolled action, from linear static model; QUF is the force-controlled action,
from linear static model; and is the strength-reduction factor from the appropriate material-specific code.
m Q CE ≥ Q UD deflection-controlled criteria
Q CL ≥ Q UF force-controlled criteria
The use of the NSP is not restricted by either irregularity or DCR limitations. The actions applied to the structure for the NSP are the same for
calculating displacement-controlled and force-controlled responses; however,
the force-displacement behavior of all structural components must be explicitly
modeled, including connections wherever they are weaker or less ductile than
the connected components or the flexibility of the connection produces more
than a 10% change in the connection forces or deformations. Tabulated plastic
rotation angles, as in ASCE 41-06, are used in the NSP to determine the
Dynamic Increase Factors; the deformation limits determine compliance for
deformation-controlled actions, and the lower bound estimate of the component
strength determines compliance for force-controlled actions.
REFERENCES
261
While the UFC relies on the best information available, based on seismic
research, the justification for the specified factors and procedures requires rigorous experimental and analytical study. Furthermore, the performance of actual structural details for the different structural materials and systems requires
extensive analytical research and development. Although simplified procedures
must conservatively represent inelastic material properties and large displacement mechanics, there is little justification for overly precise procedures to approximate the effects of dynamics and nonlinear behavior. Furthermore, the correlation between the simplified analytical representation of structural behavior
and the actual performance of structural components undergoing large deformations requires significant testing and analysis. Even the use of conventional
understrength factors, factors, may be questioned for the regime in which materials will be severely distorted.
REFERENCES
American Society of Civil Engineers. 2007. Seismic Rehabilitation of Existing Buildings
(ASCE 41-06). New York: American Society of Civil Engineers.
ASTM. 2003. Standard Practice for Specifying an Equivalent 3-Second Duration Design
Loading for Blast Resistant Glazing Fabricated with Laminated Glass (ASTM F2248).
West Conshohocken, PA: ASTM International.
ASTM. 2007. Standard Practice for Determining Load Resistance of Glass in Buildings
(ASTM E1300). West Conshohocken, PA: ASTM International.
Biggs J. M. 1964. Introduction to Structural Dynamics. New York: McGraw-Hill Book
Company.
British Standards Institute. 2006. Eurocode 1: Actions on Structures—Part 1–7: General
Actions—Accidental Actions (EN 1991-1-7:2006). London: British Standards Institute.
Federal Emergency Management Agency. 2002. World Trade Center Building Performance Study: Data Collection, Preliminary Observations and Recommendations
(FEMA 403). Washington, DC: Federal Emergency Management Agency, Department
of Homeland Security.
Hyde D. W. 1992. CONWEP, Conventional Weapons Effects (USAEWES/SS-R).
Vicksburg, MS: U.S. Army Engineer Waterways Experiment Station.
International Code Council. 2008. Supplement to the International Building Code.
Washington, DC: International Code Council.
Kingery C. N. and G. Bulmash. 1984. Airblast Parameters from TNT Spherical Air
Burst and Hemispherical Surface Burst (ARBRL-TR-02555, AD B082713). Aberdeen
Proving Ground, MD: U.S. Armament Research and Development Center, Ballistic
Research Laboratory.
New York City Department of Buildings. 2008. Supplement to the Building Code of
the City of New York, Local Law No 76 Effective December 6, 1968. New York:
Department of Buildings.
Oswald C. J. 2005. Component Explosive Damage Assessment Workbook (CEDAW)
Methodology Manual V1.0 (02-0752-001), San Antonio, TX: Baker Engineering and
Risk Consultants, Inc.
262
STRUCTURAL SYSTEMS DESIGN
Protective Design Center. 2006. Methodology Manual for the Single-Degree-of-Freedom
Blast Effects Design Spreadsheets (SBEDS) (PDC-TR-06-01), Omaha, NE: U.S. Army
Corps of Engineers, Protective Design Center.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (UFC 3-340-01). Washington, DC: Department of
Defense, Unified Facilities Criteria Program.
. 2008a. Design of Buildings to Resist Progressive Collapse (UFC 4-023-03).
Washington, DC: Department of Defense, Unified Facilities Criteria Program.
. 2008b. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: Department of Defense, Unified Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
10
Building Envelope and Glazing
Eve Hinman and Christopher Arnold
For an explosion outside a building, the exterior envelope is the critical line of defense that separates the people, operations, and contents inside the building from
the air-blast effects outside the building. Unfortunately, for most standard building types, the building envelope is ill-suited to resisting air-blast loads, which are
enormous and which act directly on the surface area of the exterior envelope in
the out-of-plane direction.
In this chapter, some of the key concepts needed to optimize the design of
the exterior envelope to mitigate explosion effects are presented. A variety of
wall and window types are considered, as well as doors and louvers. Roof and
foundation systems are also addressed.
10.1 DESIGN INTENT
The primary design objective for the exterior envelope of most standard building
types, is to mitigate the hazard of flying debris generated by failed exterior walls,
windows, and other components, to reduce casualties and business disruption and
facilitate rescue and evacuation efforts.
Other objectives are to design the exterior envelope to:
r Fail in a way that does not initiate progressive collapse
r Keep the air blast outside the building to the best of its ability
10.1.1 Life Safety
Impact of flying debris entering a building causes laceration and blunt trauma
injuries, and potentially fatalities. Once the exterior envelope is pierced and the
air blast enters the building, injuries such as eardrum rupture, lung collapse, and
blunt trauma and concussion due to being thrown against objects are the dominant injury modes. Impacts due to internal nonstructural damage to partitions
and ceiling, light fixture, and equipment components and falling furniture are
also a hazard.
263
264
BUILDING ENVELOPE AND GLAZING
Although it is cost-prohibitive to eliminate all these hazards, it is possible to
reduce the risk to a level that is acceptable through informed design decisions.
10.1.2 Emergency Egress and Facilitating Search and Rescue
The building exterior envelope design may be optimized to make it easier for
people to get out and for emergency responders to enter safely.
After an explosive event, the debris created by the failed exterior envelope
can exacerbate casualties by reducing the efficiency of evacuation and rescue
operations. Some examples of this are:
r
r
r
r
Internal debris can cause occupants to become trapped in rubble piles.
Exterior debris can block emergency exits.
Debris along egress paths can cause delays and additional injuries.
Exterior falling fragments post-event trigger safety precautions for emergency operations, which can significantly delay rescue, thereby putting the
injured at greater risk.
r Toxic dust or fumes generated by debris may have long-term health ramifications.
10.1.3 Critical Functions (Protecting Equipment and Business Processes)
Even at low pressures, air blast that enters the building can cause extensive nonstructural damage to interior partitions and suspended ceilings. The resulting
internal disarray can cause extensive business disruption. Even if the damages
are repairable, total cost required to return the building to pre-event functionality
may exceed the replacement cost of the building.
Sensitive mechanical parts of critical equipment can be disabled by air blast
or flying debris that enters the building through the exterior envelope. An interior
baffle wall may be used to dissipate the energy of the air blast by forcing it to
change direction two or more times before it reaches the equipment. Some other
options are to use storm louvers or forced-entry-resistant louvers, which have
a chevron cross section which are able to perform a similar function. Louver
attachments into the structure are to be designed to resist the capacity of the
louver to ensure that the louver fails before the connections. In some cases, a
catch system using a grate securely fastened behind the louver may be used to
prevent impact into delicate equipment components.
Critical equipment placed on the roof will be less vulnerable to a ground level
explosion than equipment at or below street level. For roof equipment, a hardened wall with roof may be used for protection. However, for this solution, note
that if the wall fails, the impact of debris may have a more detrimental effect than
using lightweight housing that is not blast-resistant.
It is recommended that the transformer be placed away from vulnerable areas,
for instance in a parking lot, near the loading dock, or under the sidewalk. Inside
DESIGN APPROACH
265
the building in a protected location is preferred. The switchgear room is to be
separated to the extent practical from the transformer(s), the generators, and any
other backup systems, such as an uninterruptible power supply (UPS) or batteries, to improve the chances of retaining at least emergency functions in the event
of an explosive attack. In addition to separation, redundancy of critical systems
is highly recommended for higher-risk buildings. An example of this is placing
one set of generators on one side of the building, and another set on the other
side of the building to create redundancy.
10.2 DESIGN APPROACH
For buildings with a low to moderate risk of explosive attack, the design
approach is to first design the building exterior envelope for conventional loads,
and then to evaluate the response to explosive loads. If the design team is
properly consulted when key decisions are made during the concept phase, then
the modifications required to meet explosive load requirements can be addressed
efficiently. This approach ensures that the design meets all the requirements for
gravity and natural hazards, in addition to air-blast effects, and is effective for
buildings where explosion effects are a secondary concern.
Take note that some features that are beneficial for blast resistance diminish
performance for other loads, if not properly integrated. For example, using a
plastic coating that emits toxic fumes in the event of a fire may not be a good
solution for enhancing the blast performance of an exterior wall, since fire is a
more common hazard than terrorist attack.
Another example that presents a potential conflict between the seismic design
and blast design is that increased mass generally increases the design forces,
whereas for explosion loads, mass generally improves response. This is because
of the highly impulsive nature of explosive loads. Heavy structural components
have longer periods of vibration so that, by the time the mass, M, is mobilized and
the structural component begins to move with a velocity, V, the load or impulse,
I, has been removed. In mathematical terms, the kinetic energy, KE = MV 2 /2, or
momentum, MV, imparted to the structural component is directly proportional
to impulse and inversely proportional to the mass: i.e., KE = MV 2 /2 = (Mv)2 /
(2M) = I 2 /(2m). In this situation, careful coordination between the blast consultant and the structural engineer is needed to provide an optimized response that
accounts for both loading cases.
Good antiterrorism design is a multidisciplinary effort, requiring the concerted efforts of the architect, structural engineer, security professional, and the
other design team members. For instance the passive protection provided by
effective building design needs to be balanced with the active or operational
security measures implemented (e.g., guards, cameras, dogs, sensors, magnetometers). It is critical that these two sets of measures complement and support
one another. In particular, the size weapon that is considered at the perimeter
and at the exterior envelope is a function of the size weapon that is able to be
266
BUILDING ENVELOPE AND GLAZING
easily detected through operational security measures. Like protective design,
security needs to be implemented early in the design to be effective. Lines
of sight; lighting and placement of cameras along the exterior envelope; and
location, number and type of entrances provided are examples of security issues
to be addressed early in design with the architect, since they potentially affect
the design threat to be used.
The envelope system is to be designed by keeping in mind the concepts of
balanced or capacity design (see Section 10.2.3) and ductile response. Ductile
systems are able to absorb energy through inelastic flexural deformation. Systems that fail in a brittle manner through shear are to be avoided. By permitting
limited exterior envelope damage that does not significantly increase the hazard to the majority of the occupants, it is possible to design more cost-effective
systems that absorb energy through deformation, and transmit lower forces into
the connections and supporting structure, thus reducing the potential for more
serious structural failures.
Perhaps the simplest analytical method is to use a pressure-impulse, or P-I,
diagram, in which various damage levels for a component with a given resistance are expressed as curves, and the axes are peak pressure and impulse. By
plotting the pressure and impulse acting on the system, you can estimate the
level of damage to the exterior envelope component and compare it with allowable limits. These curves may be based on explosive tests or on the output from
a semi-empirical or a single-degree-of-freedom system. The limitation of these
tools is that they are based on constant parameters for the component under consideration. If the component that is being investigated differs in any way, some
extrapolation or interpolation between charts is needed.
Typically, simplified or single-degree-of-freedom (SDOF) methods used for
the design of structural components may be used for the exterior envelope (Unified Facilities Criteria Program 2008). For these systems, the component timedependent peak deflection, X(t), is evaluated by modeling the component as a
lumped mass, M, so that the inertia force is MA(t) where A(t) is the acceleration
or second derivative of X(t). The resistance is provided by a spring, R(x), which
has a linear elastic flexural stiffness, K, and a ultimate flexural resistance of Ru .
The loading is modeled as a time-dependent function, F(t). For a system initially
at rest, the resulting equation of motion is:
M A(t) + R(x) = P(t)
where: R(x) = KX(t) when X(t) is less than or equal to the maximum elastic limit
displacement Xe and
R(x) = Ru when X (t) > X e
In these expressions, M and R, and P are “equivalent” values which represent the portion of the mass, resistance, and loading that is participating in
the displacement of the member, and which provide a natural frequency and
DESIGN APPROACH
F(t)
267
R
RU
KLMM
K
1
Xe
R(x)
X
Xm
F
Fl
td
f
Figure 10.1 Graphical Representation of an SDOF System
displacement that are the same as that of the actual system. Figure 10.1 provides
a graphical representation of an SDOF system.
These equations may be easily solved by using numerical methods. The
resulting maximum displacement is typically evaluated by comparing the ratio
of the maximum displacement Xm and the maximum elastic limit displacement,
or ductility Xm/Xe. Another measure is the support rotation, which is defined
as the angle between the undeflected member and the line connecting the
support point to the point of peak deflection. For example, for a uniformly
loaded, simply supported beam of length L the support rotation is equal to the
tangent of 2Xm/L. Figure 10.2 illustrates the concepts of displacement ductility
and rotation. Typically, the maximum allowable ductility may vary from 1
for a brittle material like monolithic glass to about 10 for ductile systems like
steel. For heavily damaged systems, a ductility of up to 30 may be allowed.
Permitted ductilities are generally about 2 degrees for most systems, but may be
permitted to be 4 degrees or more for highly ductile structures where extensive
but low-hazard damage is permitted.
Other models that may be used include two or more degrees-of-freedom systems or piecewise linear continuum models. Two-degrees-of-freedom models
have been used with some success, in particular for a beam and girder where
a portion of the load is transmitted directly to the girder from the beam, and
where the beam experiences rigid body as well as deformational displacements.
Piecewise linear systems are of particular interest because they are able to represent behavior more accurately for bilinear or trilinear resistance functions and
loading functions. This type of analysis was not a practical solution until the
recent advent of powerful computational tools such as MATLAB (MathWorks
2006) or Mathcad (MathSoft 2001).
However, for complex exterior envelope systems or air-blast loading scenarios
or systems using innovative materials or energy absorption methods, simplified
268
BUILDING ENVELOPE AND GLAZING
R
RU
K
X
Xm
Xe
L
Xm = Peak Displacement
Plastic Hinge
Figure 10.2 Graphical Representation of the Concepts of Displacement Ductility and
Rotation
analysis is not sufficient. In these cases, the engineer needs to use finite-element
numerical time integration techniques and/or explosive testing to verify response.
Again, nonlinear dynamic time history analysis is required to adequately model
the system. Material models used for these programs are generally more sophisticated than the linear elastic perfectly plastic models used for simplified analysis. In this case, the material models provide a more realistic representation
for large deformation. Examples of computer codes on the market that are capable of performing these types of computations include SAP2000/NonLinear
(Computers and Structures 1990) and LS-DYNA (Livermore Software Technology Corp. 1999). There are also codes available to model the expansion of an air
blast and its reflection against the structure surface, which may be used to generate the loads applied to the model using a software product such as CONWEP
(Hyde 1988) or ATBLAST, a software product prepared by Applied Research
Associates.
The time and cost required to perform the analysis cannot be ignored in choosing computational methods. Because the design process may require iteration,
the cost of analysis must be justified in terms of benefits to the project and increased confidence in the reliability of the results. In some cases, a simplified
approach will be used for the preliminary design and a more sophisticated approach, using finite elements and/or explosive testing, may be used for the final
verification of the design.
If testing has been performed for a similar system, it is necessary to obtain the
testing report and verify that the assumptions and findings are consistent with
the project requirements, prior to accepting testing in lieu of detailed analysis.
In particular, check to make sure that the support conditions for the member are
DESIGN APPROACH
269
accurately represented. Also check to see that both the pressure and impulse are
at least as large as what the design requires. If only the pressure or the impulse
meets criteria, some additional calculations will be needed to verify response.
It is recommended that the testing be supplemented with pre- and post-analysis,
using an advanced method such as finite elements.
Often, only a proof of concept design is provided by the blast engineer during
design. The final blast design is the responsibility of the subcontractor implementing the design. A detailed set of specifications is strongly recommended
to ensure that the design intent is met. Some of the requirements to consider,
including in the specifications for the exterior facade subcontractors, are:
r Provide performance-based design requirements for response.
r Put all the blast requirements in a single location in the specifications, rather
than sprinkling them through the individual sections. This makes it easier
for everyone to keep track of these special requirements.
r Require the blast engineer’s approval of the subcontractor based on successful performance on similar projects.
r Require the subcontractor to submit blast calculations for review.
10.2.1 Response Criteria
Levels of damage to the exterior envelope may be described by these terms:
minor, moderate, major, catastrophic. Qualitative descriptions of each damage
level are given below as they pertain to the exterior envelope. Note that all these
descriptions imply that there is little or no structural damage (i.e., all the damage
is limited to the exterior envelope).
Minor: There is little if any damage. The exterior envelope may sustain some
damage, but virtually no failure. Damage may consist of cracked windows,
failed sunshades still attached to the building, or permanent deformation
of walls and window frames. Injuries may be expected, and fatalities are
possible but unlikely.
Moderate: Failure of the exterior envelope is confined to a localized area and
is usually repairable. Hazardous window failure is limited to one face of
the building. Injuries and possible fatalities are expected.
Major: Extensive exterior envelope damage and some roof damage are expected. In this case, extensive fatalities are expected. Damage is usually
not repairable.
Catastrophic: In this case, all or the majority of the exterior envelope has been
removed.
Generally, moderate damage for the design basis threat is a reasonable goal
for new construction to meet life-safety objectives. For buildings that need to
270
BUILDING ENVELOPE AND GLAZING
remain operational during and after an event, or are designated as high risk,
minor damage may be the more appropriate damage level.
Quantitative performance is usually evaluated by comparing the ductility (i.e.,
the peak displacement divided by the elastic limit displacement) and/or support
rotation (the angle between the support and the point of peak deflection) to empirically established maximum values that have been established by the military
through explosive testing. Maximum permissible values vary depending on the
material and the acceptable damage level. Some criteria documents do provide
the design values that need to be met.
10.2.2 Static versus Dynamic
A dynamic nonlinear approach is more likely to provide a design that meets the
design constraints of the project than a static approach. Elastic static calculations are likely to give overly conservative design solutions if the peak pressure
is considered without the effect of load duration. By using dynamic calculations
instead of static, we are able to account for the very short duration of the loading. In addition, the inertial effect included in dynamic computations greatly improves response. This is because by the time the mass is mobilized, the loading
is greatly diminished, enhancing response. Furthermore, by accepting that damage occurs, we are able to account for the energy absorption of ductile systems
that occurs through plastic deformation. Finally, because the loading is so rapid,
we are able to engage the enhanced material strength that often occurs with very
high strain rates.
10.2.3 Balanced Design
Balanced or capacity design philosophy for a building system refers to designs
that are controlled such that the failure of the weakest component in the system
results in the least destruction. For a window system, for instance, it refers to the
window glass failing at pressure levels that do not exceed those of the frame, anchorage, and supporting wall system. If the glass is stronger than the supporting
members, then the window is likely to fail with the whole panel entering into
the building as a single unit, possibly with the frame, anchorage, and the wall
attached. This failure mode is considered more hazardous than one in which
the glass fragments enter the building, provided that the fragments are designed
to minimize injuries. By means of a damage-limiting approach, the damage sequence and extent of damage are controlled.
10.2.4 Load Path
Load path in the exterior envelope generally refers to the transmission of load
from the member that presents the most surface area to the exterior (i.e., glass
pane, panel, wall, slab) into the building structure. Connections and reaction
forces, two of the more nontrivial parts of the path, are discussed below.
DESIGN APPROACH
Connections
271
For the exterior facade, the connections include:
r The connection between the window glass and the window mullion, transom or frame
r The connection of the window mullion or transom and the window frame
r The connection between the window frame and the wall
r The connection between the wall and the structural frame
Connections are typically the least ductile elements in the exterior envelope
system and the most prone to failure if not properly designed. The connections
are required to enable the supporting elements to reach their ultimate flexural
capacity without failure. They also need to remain intact for the loads imposed
by large inelastic deformations of the supporting members, which in turn can
cause significant deformation of the connections.
Fasteners holding the window frame in the wall tend to be placed into tension
and shear as the window undergoes deformation. The wet glazing are put into
shear as a laminated glass pane begins to act as a membrane. As the mullion of a
unitized system deforms, stresses are placed on the connections holding the two
halves of the mullion together.
The failure envelope of connections respondingto load combinations such as
tension and shear need to be considered as part of the design.
Although it seems as if window frame connections should be routinely designed to carry the ultimate loads transmitted by the window pane, this is not the
case. The window frame is designed to carry the design loads only. The capacity
of the glass may be significantly higher than the design loads.
Reaction Loads (Ru versus Applied Load) Reaction loads may conservatively
be approximated by the static ultimate capacity of the loading element, and by
setting the load and strength factors for the connection design to 1.0. The load
factor is set to 1 because we are assuming that the supported member has been
pushed to its capacity. The reaction load cannot be higher than the capacity of the
member. Similarly, the strength factor is set to 1, again in recognition of the fact
that we are letting the supporting member reach its ultimate capacity. Although
this may not seem conservative, it is, because we neglected the mitigating effects
of ductility and inertia on the system.
This approach is consistent with capacity or balanced design principles and
is the best and most reliable. However, it can lead to designs that are not constructible or affordable. If this is the case there are several approaches which
may be considered:
r Use the average applied load instead of the peak load, because the edge
load along a panel transfers a load that varies with its peak value at the
center point and decreases towards the ends. Often the peak value is given,
but the average load along the length is less and is more representative of
the loading on the anchors.
272
BUILDING ENVELOPE AND GLAZING
r Increase the strength factor for flexural or tensile response, to account for
strain-hardening effects. For impulsively applied loads, the strength of the
material is increased. This effect is referred to as strain hardening.
r Permit the loss of a secondary member (for instance, permit the loss of a
beam but not a girder). By enabling the major element in the load path to
be designed to the capacity of the elements it supports, we are improving
performance while maintaining the cost-effectiveness of the design. This is
done often for roof systems where the floor slab capacity load is greatly in
excess of the design loads.
r Use an element with a lower ultimate capacity. This is a practical and effective tool for systems where there is significant ductility. For instance, a thin
precast panel will have a lower ultimate resistance than a thicker panel and
can be designed to perform well in ductility.
10.3 FENESTRATION
Windows, once the sole responsibility of the architect, become a structural issue
once explosive effects are taken into consideration. Windows designed to mitigate the effects of explosions are first to be designed to resist conventional loads,
and then to be checked for explosive load effects and balanced design.
Even if the building remains standing and no structural damage occurs, extensive injuries can occur due to nonstructural damages. Windows are typically
the most vulnerable portion of any building. Although it may be impractical to
design all the windows to resist a large-scale explosive attack, it is desirable to
limit the amount of hazardous glass breakage to reduce the injuries. Typical glass
windows break at low pressure and impulse levels, and the shards created by broken windows are responsible for many of the injuries incurred due to large-scale
explosive attack.
Designing windows to provide protection against the effects of explosions can
be effective in reducing the glass laceration injuries. Glass lacerations are caused
when the shards of glass created when a window is broken are propelled into the
building and penetrate the skin. This is perhaps the most common type of injury
that results from large-scale explosion incidents and may occur hundreds or even
thousands of feet from the source in urban areas, where the air blast attenuates
more slowly as it propagates through the street “canyons.”
Window protection should be evaluated on a case-by-case basis by a qualified
protective design consultant, to develop a solution that meets established objectives. In the sections below, a number of generic recommendations are given for
the design of the window systems to reduce injuries to building occupants.
To limit glass laceration injuries, there are several approaches that can be
taken. One approach is to reduce the number and size of windows, which
potentially will reduce the air-blast and glass shards entering the building,
FENESTRATION
273
thus reducing the interior damage and injuries. Specific examples of how to
incorporate these ideas into the design of a new building include:
r Limiting the number of windows on the lower floors where the pressures
are higher due to an external explosive threat
r Using an internal atrium design with windows facing inward toward the
atrium, rather than outward toward the facility perimeter
r Windows, which are close to the ceiling, above the heads of the occupants
r Angling the windows away from critical weapon locations to reduce the
pressure levels
10.3.1 Glass
Glass is often the weakest part of a building, breaking at low pressures compared to other components such as the floors, walls, or columns. Past incidents
have shown that glass breakage and associated injuries may extend many thousands of feet in large external explosions. High-velocity glass fragments have
been shown to be a major contributor to injuries in such incidents. For incidents
within downtown city areas, falling glass poses a major hazard to passersby and
prolongs post-incident rescue and cleanup efforts by leaving tons of glass debris
on the street.
As part of the damage-limiting approach, glass failure is not quantified in
terms of whether breakage occurs or not, but rather by the hazard it causes to the
occupants. The glass performance condition is defined based on empirical data
from explosive tests performed in a cubical space with a 10-foot dimension. The
performance condition ranges from 1, which corresponds to no glass breakage,
to 5, which corresponds to hazardous flying debris at a distance of 10 feet from
the window.
Design criteria established by the Interagency Security Council, which governs the antiterrorism efforts for most agencies, require performance condition
3b for buildings that are at moderate risk of attack. At this protection level, the
window breaks, and fragments fly into the building but land within 10 feet of the
window. The design goal for moderate-risk buildings is to achieve a performance
level of less than 4 for 90% of the windows. Protection level 4 means that glass
fragments will penetrate a witness panel located 10 feet behind the window for a
height of two feet or less.
The preferred solution for new construction is to use laminated glass with
structural silicone sealant (i.e., wet glazing) around the inside perimeter. The
lamination holds the shards of glass together in explosive events, reducing its
potential to cause laceration injuries. The structural sealant helps to hold the
pane in the frame for higher loads. For insulated units, only the inner pane needs
to be laminated because the inner pane will provide protection against the hazard
of shards generated by the outer pane.
274
BUILDING ENVELOPE AND GLAZING
Annealed glass has a breaking strength that is about one-half that of heatstrengthened glass and about one-fourth that of tempered glass (Amstock 1997),
thus reducing the loads transmitted to the supporting frame and walls. To reduce
reaction loads, it is sometimes advantageous to consider annealed glass. The
preferred interlayer thickness is 60 mil (0.06 in) unless otherwise specified by
the criteria. At a minimum, 30 mil is to be used.
To make sure that the components supporting the glass are stronger than the
glass itself, we specify a window breakage strength that is high compared to
what is used in conventional design. The breakage strength in window design
may be specified as a probability of the number of windows expected to break
at that load. For instance, when designing for conventional loads, it is typical
to use a breakage pressure corresponding to a probability of 8 breaks out of
1,000. Where there is a high probability of extensive glass breakage, as in an
explosion incident, a pressure corresponding to 750 breaks out of 1,000 is used.
Glass breakage strengths may be obtained from window manufacturers.
Smaller glass panes generally have higher capacities than larger panes. Consequently, smaller panes can cause significantly higher loads to be transmitted
to the frames than larger panes. As a result, in blast-mitigating design, we avoid
small panes. Also, since every size pane has a difference capacity, it is desirable
to standardize the design as much as is practical.
There are several government-sponsored software products available for evaluating the response of window glass for use on federal projects, including HAZL
(U.S. Army Engineer Research and Development Center 2001), WINGARD,
and WINLAC. These codes are made available to government contractors who
have government projects requiring this type of analysis. One approach that
is publicly available is described in the criteria by the U.S. military (Unified
Facilities Criteria Program 2007), using ASTM Standards (ASTM 2003,
2004a). For those who are designing unique systems for specific threats not
covered in criteria documents, it is recommended that the system be tested
(ASTM 2001).
Glass block is generally not recommended because of the heavy projectiles
these walls may create due to failure at the mortar lines. However, blast-rated
glass-block products are available in which the glass blocks are framed by a
steel grate system or other method.
The way in which the glass is supported into the exterior envelope can have
an effect on the severity of the glass hazard. A brief summary of various types of
supports is given below.
Punched Windows: Punched windows (see Figure 10.3) often can be designed
efficiently to be less vulnerable than other window types because the frame is
attached directly into the wall, which is generally of more robust construction.
Strip Windows: Strip or ribbon windows (see Figure 10.4) generally consist
of alternating bands of glazing and opaque material constructed using precast
panels, poured concrete, or insulated metal panels. The opaque area conceals
the floor structure and is referred to as a “spandrel.” Ribbon windows have thin
vertical metal framing separating the individual panes, and are supported by
FENESTRATION
275
Figure 10.3 Example of “Punched” Windows
the wall at the top and bottom. For this type of exterior facade, the spandrels
are attached to each floor level instead. The system is very economical and a
common facade type. To resist air-blast loads, steel angle kicker braces are often
needed to laterally support the bottom of the spandrel panel.
Glass Curtain Wall: For this glazing type, a significant part of the building exterior is covered with windows supported by aluminum or possibly steel framing.
Figure 10.5 shows three variations of curtain wall system.
Figure 10.4 Example of Ribbon Windows, Showing Spandrel
(a)
(b)
(c)
Figure 10.5 Types of Metal and Glass Curtain Walls. (a) Curtain wall framing expressing main building structure. (b) Uniform curtain wall framing over the fac¸ade. (c) curtain
wall framing draped over rectilinear building forms.
FENESTRATION
277
Figure 10.6 Example of Point and Cable Supported Curtain Wall
This support type has an increased risk of hazardous failure as compared with
punched windows because strip window systems are supported by wall on two
sides instead of four sides.. However, it has more flexibility, which allows it to
deform significantly without failing.
Point and Cable Supported: In point and cable supported systems (see
Figure 10.6), each window pane is connected at points near its corners. The
bracket that supports the glass often has multiple arms, forming a spider, so that
it can attach near the corner of several adjacent panes. The glass is drilled for a
threaded connector to pass through, engaging the interlayer to secure the pane
during an explosion. The panes are typically connected using a clear or translucent polymer material instead of the metal framing (referred to as “butt glazed”).
Spandrels The term “spandrel” is used to describe the non-vision panels used
above and/or beneath the window or at the floor levels (see also discussion regarding “strip windows” above). They are used to hide from view the structure
or mechanical equipment behind. Sometimes a “shadow box” is used behind the
spandrel, which is a metal pan designed to catch the glass fragments and keep
the explosive loads from entering the building.
Operable versus Inoperable Although operable windows can be designed to
meet modest explosion requirements, keep in mind that they will not keep air
blast out of the building if they are open.
Inoperable window solutions have potential to be more reliable for protection
during air-blast events, because occupants cannot void the windows’ protective
function by opening them. However, there are operable window solutions that
are conceptually viable. For instance, if a window is designed to open outward
278
BUILDING ENVELOPE AND GLAZING
about a horizontal hinge at the sill, the window will tend to slam shut in an
exterior explosion. If this type of design is used, the governing design parameter
may be the capacity of the hinges and/or hardware.
The controlling design component for sliding operable windows (and glazed
doors) often is the supporting track. If possible, the track should be embedded in
the supporting structure to provide added out-of-plane support to the system.
Skylights/Courtyards Skylights in roof systems have the advantage of not being subject to reflected pressures fromexterior explosions at ground level. Also,
skylights are required by building code to be designed using laminated glass,
which is preferred for explosion-mitigating design. However, they do create a
falling fragment hazard. Therefore, these should be designed with a catch system beneath, or designed to remain in the frame for the design air-blast load.
Ideally, skylights should be placed as far from the weapon as possible, to keep
the pressures low.
For an atrium with an open courtyard in the center, the windows are protected
by the building on all sides, and the glass will be subject to indirect air-blast
pressures in the event of an explosion outside the building exterior. Because atria
are protected on all sides, they may be designed for the reduced pressures from an
exterior threat. These reductions in pressures can help significantly in reducing
the design requirements imposed by explosion loads.
Doors Glazed doors and glass panes above a door are to be designed using
the same methods as for windows. Glazed doors tend to use tempered glass for
safety, in case of accidental breakage due to impact. As a result, they are highstrength and can complicate the design of the supporting system
10.3.2 Mullions/Transoms
The vertical frame members connecting adjoining windows are referred to as
mullions. Horizontal members are termed transoms. These members may be
designed in two ways. Either a static approach may be used, whereby the
breaking strength of the window glass is applied to the mullion, or a dynamic
load may be applied, using the peak pressure and impulse values. A static approach may be overly conservative because the designer using this approach
assumes that the peak edge forces are sustained indefinitely, whereas in an
explosion the forces are of very short duration, and the supporting structure
has insufficient time to fully respond to the impulsive forces. Using this approach, the mullion can become very deep and heavy, driving up the weight
and cost of the window system. Figure 10.7 shows heavy mullions in an older
building.
Sometimes cables or steel bars or tubes attached to the supporting structure are
placed behind the glass to prevent the laminated glass or glass with anti-shatter
film from entering the interior (Crawfordk, Lan and Dunn 2006).
FENESTRATION
279
Figure 10.7 Example of Mullions in an Older Building
10.3.3 Frame and Anchorage
The window frames need to retain the glass so that the entire pane does not
become a single large unit of flying debris.
To retain the glass in the frame, a minimum of a 1/4 -inch bead of structural
sealant (e.g., silicone) is used around the inner perimeter of the window. For
designs where a 60-mil polyvinyl butyral (PVB) interlayer is used, the silicone
sealant should be designed to resist the shear forces caused by the membrane action forces using the ultimate tensile capacity of the PVB material. These membrane forces can significantly increase the reaction loads into the framing system,
increasing the design requirements for the entire window system. The allowable
tensile strength of the silicone sealant should be at least 20 psi. Also, the window
bite (i.e., the depth of window captured by the frame) needs to be at least 1/2 -inch.
For large windows, such as in the lobby area, the bite required for conventional
loads may well exceed the 1/2 -inch minimum, and additional silicone sealant may
be needed to avoid failure caused by the entire glass pane exiting the frame.
280
BUILDING ENVELOPE AND GLAZING
Frame and anchorage design is performed by applying the reaction forces
from the window at breakage to the frame and the fasteners. In most conventionally designed buildings, the frames will be aluminum. In some applications,
where the windows are designed to resist high pressures and for long spans, steel
bar inserts, cable inserts, or built-up steel frames may be used.
For reinforced concrete construction designed to resist high-pressure loads,
as is typical for embassy construction, anchorage of the steel window frames is
provided by steel studs welded to a steel base plate. For this type of construction,
the frame is typically constructed using a steel stop at the interior face and an
angle with an exposed face at the exterior face. The frame is attached to the base
plate using high-strength fasteners. Coordination between the wall and window
subcontractors is required to ensure that the fastener locations are spaced so that
they fit between the rebar in the wall.
For masonry walls, flat metal straps embedded in the mortar with adequate
development length are recommended for anchoring the window into the wall.
10.3.4 Supporting Structure
It is inconsistent with balanced design and is potentially highly hazardous to have
a wall system that is weaker than the windows it is supporting.
Anchoring window/wall systems into columns is generally discouraged because it increases the tributary area of lateral load that is transferred into these
critical members, and may cause failure or instability.
Some window/wall designs will require additional lateral support. For instance, when the supporting walls are acting largely as cantilevers, such as for
windows placed high on a wall, they might need to be supported with vertical
braces spanning the clear floor height using for instance steel tubes to provide
additional stiffness. For punched wall systems with narrow pilasters between
them, vertical braces may also be needed. For lighter wall systems such as metal
stud systems, suitable reinforcements such as back-to-back double studs framing
the window are recommended.
The balanced design approach is particularly challenging in the design of
ballistic-resistant and forced-entry-resistant windows, which consist of one or
more inches of glass and polycarbonate. In this situation, the window may be
framed directly to the structure using steel framing to transfer loads and not into
the wall.
10.3.5 Other Penetrations
Similar to windows, other penetrations are designed to improve the safety of the
occupants. Rather than explicitly designing to resist air-blast pressures, the focus
is on reducing the hazard presented by the failure.
Doors Doors are handled differently in different criteria documents. Most criteria documents neglect the response of door systems. This may be for several
reasons. Doors that are capable of resisting air-blast loads can be very expensive.
EXTERIOR WALLS
281
Also, doors are typically in transitory areas where people do not stay for very
long. Some concepts for increasing the inherent protection offered by doors are
as follows:
r Use double steel doors with internal cross braces.
r Orient doors to open outward so that they bear against the jamb during the
positive pressure loading phase.
r Fill the jambs with concrete to increase their strength.
r Increase the number of fasteners used to connect the door into the wall
system.
Revolving doors are required by code to use tempered or laminated glass, to
prevent impact breakage during normal use. It is not typical to explicitly design
these doors to resist the threat, but some precautions may be taken to reduce the
hazard, such as increasing the bite or using silicone sealant between the glass
and frame.
Louvers Louvers are another type of opening to consider. These components
should be designed with connections that are able to resist the flexural capacity
of the louver. A catch system, consisting of a well-anchored steel grate behind
the louver, is another approach.
Air intakes that are at or close to the ground level should always have grates so
that weapons cannot be lobbed into them. Also consider using a sloped grating
(at an angle of at least 45 degrees from the horizontal) so that a potential weapon
can roll off prior to detonation.
Blowout Panels Blowout panels are designed to disintegrate or break free at
relatively low impulses, to limit the buildup of expanding hot gasses in an interior
space. This type of wall fails easily and will act to vent the explosive forces outside the building, helping to protect vertically and horizontally adjacent spaces.
Historically, blowout panels also are used when designing for an accidental
explosion inside a building. For deliberate explosions, the approach is somewhat different. Blowout panels typically are used when an internal space, such
as a mailroom designed to resist a package bomb, is vulnerable to explosive attack. In these situations, mailrooms are placed adjacent to the exterior wall of
the building, and the exterior wall is designed using lightweight construction using, for instance, metal studs. Alternatively, conventional construction, without
consideration to explosion loads, may be used.
10.4 EXTERIOR WALLS
Exterior walls often are designed to resist reflection-magnified pressures from
an external explosive threat located directly beyond the secured perimeter line.
One important objective of the design is to detail these members to fail in a
282
BUILDING ENVELOPE AND GLAZING
ductile mode such as flexure, rather than a brittle mode such as shear. The walls
also need to be able to resist the ultimate loads transmitted by the windows and
doors.
It may be cost-efficient to consider the reduction in pressure with height due to
the increase in distance and the angle of incidence at the upper levels of a highrise building. Even if pressure reductions are taken into account at the upper
floors, minimum requirements such as balanced design and ductile response are
to be met, to reduce the hazard to occupants in case the actual explosion is greater
than the design threat.
Various types of wall construction are considered below.
10.4.1 Concrete Walls
Cast-in-Place Historically, the preferred material for explosion-mitigating
construction is cast-in-place reinforced concrete. This is the material that is used
for military bunkers, and the military has performed extensive research and testing on its performance. Reinforced concrete has a number of attributes that make
it an attractive construction material because of its mass, ductility, and monolithic
character.
Reinforced concrete is recommended for high-risk buildings that are vulnerable to large-scale attack, and require a high level of protection. It is also recommended for highly protected areas within buildings, such as primary egress paths
or high-occupancy areas.
Note that for reinforced concrete to respond favorably to explosion loads, it is
to be detailed in a ductile manner, such as is done in high seismic zones. Some
attributes of ductile protective design (Federal Emergency Management Agency
2003) are as follows:
r Use symmetric reinforcement on both faces, so that load reversals may be
accommodated.
r Span the wall from floor to floor, rather than from column to column, to
reduce the chances of building collapse.
r Stagger splices away from high-stress areas, to maintain the monolithic
character of the material.
r Space reinforcing bars no more than one wall thickness apart, but no less
than one-half the wall thickness apart, to lessen the chances of localized
brittle behavior.
r Use ductile special seismic detailing at plastic hinge locations, so that they
are able to develop the full moment of the span.
r Use full tensile development lengths to develop the ultimate flexural capacity of the section.
r For progressive collapse prevention, consider the loss of an exterior wall
that measures vertically one floor height and laterally one bay width.
EXTERIOR WALLS
283
r Use closed ties or spiral reinforcing along the entire length of beams and
columns, including connections with a minimum bend angle of 135◦ and a
spacing not exceeding d/2, to provide confinement.
Precast Panels For precast panels, consider a minimum thickness of five
inches, exclusive of reveals, with two-way reinforcing bars to reduce the ultimate
resistance, increase ductility, and reduce the chance of flying concrete fragments.
This will also reduce the reaction loads transmitted into the connections. In highpressure regions, using ribbed panels (see Figure 10.7) may be an effective way
to resist the loads. These panels need to bear against the floor diaphragms.
The following are recommendations and considerations in designing precast
elements for air-blast resistance.
Reinforcement: Two symmetric, continuous layers of two-way reinforcement
are recommended to accommodate large deformations and rebound loads. For
thin panels where it is difficult to place two layers of reinforcement, the use of
two layers of heavy wire mesh, one layer of two-way reinforcement along the
centerline, or staggered bars on either face may be considered.
If a single layer of reinforcement is used, it is critical to design the section so
that the steel yields before the concrete fails in compression to obtain a ductile
response. Enhanced protection may be provided by placing fiber-reinforced polymers, geotextile materials, Kevlar, or similar materials on the inside face to provide confinement, resistance to scabbing and spalling, and added tensile capacity.
If reinforcing bars are used, a layer of wire mesh on the interior face may help to
further restrain concrete fragments from entering the space. Closely spaced bars
also help fragment restraint, but care must be taken not to increase the ultimate
Figure 10.8 Precast Concrete Ribbed Wall Panels
284
BUILDING ENVELOPE AND GLAZING
flexural capacity too much, so as to keep the reaction loads to a reasonable level.
Note that centerline reinforcement will not work in the rebound direction due
to the failure of the tension concrete during the positive phase. If the primary
objective is to protect occupants and not passersby, then this may be acceptable.
Above major egress points, however, where debris outside the building presents
an obstacle to ingress and egress post-event, added protection is desirable.
Load-Bearing Systems: For progressive collapse resistance in load-bearing
precast systems, panels need to be designed to span over failed areas by means
of arching action, strengthened gravity connections, secondary support systems,
or other means of providing an alternate load path.
Connections: Ductile connections are recommended. The connections and
supporting structure need to be able to resist the loads transmitted by the panel
loaded to its ultimate flexural capacity, so that the system is balanced for the possibility that the actual air-blast loading is higher than the design load. Reaction
loads from the windows at ultimate capacity are to be included in the calculation
of connection design loads. Using this approach, every panel with a different
configuration will have a different set of design loads for the connections. Note
that small panels will have higher reaction loads than larger panels, using this
method, due to their increased stiffness. Standardizing the panel sizes greatly
simplifies the connection load determination.
Also, the connections need to provide sufficient lateral restraint for the panels
to accept large deformations. Depending on the design details of the connection
used, lateral restraint design will require consideration of in-plane shear, buckling, flexure, and/or tension loads in the design. Punching shear through the panel
also needs to be checked. The connection should provide a direct load path from
the panel into the supporting structure to minimize P-Delta effects.
Floor-to-floor panels with continuous, gravity-type connections directly into
the floor diaphragms are preferred. Multistory panels that are hung on the floor
diaphragms may also be considered. Connections into exterior columns or spandrel beams are discouraged, to avoid the possibility of initiating structural collapse of the exterior bay. Redundant gravity connections are strongly recommended, to prevent falling debris if a single connection fails.
Connections should be checked for rebound loads. It is conservative to use
the same load in rebound as for the inward pressure. More accurate values may
be obtained through dynamic analysis or charts provided in military handbooks
(Unified Facilities Criteria Program 2008).
Specifications: Specifications for precast elements should be in the form of
a performance requirement, with the air-blast pressures and required response
limits. The performance specifications give precast contractors more flexibility
to provide the systems with which they are most familiar. This approach requires
that the contractors have either in-house dynamic analysis capability or a relationship with a blast engineer who can work with them to customize the most
cost-effective system.
For structural precast systems, the connections are the critical issue to be
addressed. Connections need to permit the panel to reach its ultimate flexural
EXTERIOR WALLS
285
capacity. Extensive use of transverse walls or an “egg crate” type of design is
an effective way to achieve the needed lateral support. Special seismic detailing (American Concrete Institute 2008) is recommended for structural precast
connections.
Tilt-Up Construction Tilt-up construction is acceptable, provided that the connections into the structure are able to withstand the ultimate capacity of the wall
in both the inward and outward (i.e., rebound) direction.
Post-Tensioned Panels Pre-tensioned or post-tensioned construction without
significant mild steel reinforcement provides little capacity for abnormal loading
patterns and load reversals. If these systems are used, it is recommended that
mild reinforcing be added, to provide the needed ductility
10.4.2 Masonry
Reinforced Masonry For concrete masonry unit (CMU) block walls, as a minimum use 8-inch block walls, fully grouted with vertical centered reinforcing
bars placed in each void, and horizontal ladder type reinforcement at each layer.
Connections into the structure are to resist the reactions associated with the wall
loaded to its ultimate lateral capacity. For infill walls, avoid transferring loads
into the columns if they are primary load-carrying elements.
The connection details may be very difficult to construct. It will be difficult to
have all the blocks fit over the bars, near the top, and it will be difficult to provide
the required lateral restraint at the top connection. A preferred system is to have
a continuous exterior CMU wall that laterally bears against the floor system.
For infill walls, use development lengths that develop the full capacity of the
section. For increased protection, consider using 12-inch blocks with two layers
of vertical reinforcement. Also, consider using CMU block units that encourage
a homogenous response, such as an “I”-shaped unit with inner and outer faces
that are connected with a small strut.
Unreinforced Masonry Brick load-bearing walls are unable to sustain tensile
loads, so they must resist explosion loads mostly through mass to reduce the
kinetic energy that must be absorbed. As a result, thicker solid walls on the
order of approximately 18 inches are preferred and respond well at air-blast levels less than 10 psi or so. Unreinforced structural masonry is considered a very
brittle material that may generate highly hazardous flying debris in the event of
an explosion, and is to be avoided for new construction.
10.4.3 Steel
Metal Studs For metal stud systems, use metal studs back to back and mechanically attached. The two-stud system provides the benefit of lateral torsional resistance, and therefore can be more efficiently designed. To catch exterior cladding
286
BUILDING ENVELOPE AND GLAZING
fragments, attach a wire mesh, steel sheet, or fiber-reinforced polymer to the interior side of the panel. The supports of the wall should be designed to resist the
ultimate out-of-plane bending capacity load of the system.
If a single-stud system is used, deeper, thicker channels are preferred. As a
minimum, consider 16-ga systems with depth of 6 inches, and laterally braced
are preferred.
Special care is required at the connections to prevent failure prior to the stud
reaching its ultimate capacity.
Enhanced protection may be provided by placing fiber-reinforced polymers,
geotextile materials, Kevlar, or similar materials on the inside-face of the
cladding to provide confinement, fragment restraint, and added tensile capacity.
Metal Panels Non-structural metal panels may be successfully used if they
are braced behind with steel tubes or compact sections. Continuous welds are
required around the perimeter braces to resist the membrane loads. At the end
panels, the braces will need to be augmented to account for the lack of equalizing
forces.
10.4.4 Other
Appurtenances Brick veneers, bris soleil, sunshades, and other nonstructural
elements attached to the building exterior are to be avoided, to limit flying debris
and improve emergency egress by ensuring that exits remain passable. If used,
they should be designed using lightweight materials with connections designed
to resist the ultimate capacity of the appurtenance.
Exposed Structural Systems For exposed exterior frame systems, there are two
primary considerations. The first is to design the exterior columns to resist the
direct effects of the specified threats. The second is to detail the exterior frame so
that it has sufficient structural integrity to accept localized failure without initiating progressive collapse. As a rule of thumb to meet these goals, column spacing
should be limited to 30 feet, and floor heights should be limited to not greater
than 16 feet, wherever possible. It is recommended that the columns be designed
to resist the explosive loads associated with a contact charge placed at the base,
or that a column removal analysis be performed to verify that progressive collapse is not initiated.
Because columns do not have much surface area, air-blast loads on exposed
standalone columns that are not supporting adjacent wall systems tend to be
mitigated by “clearing-time effects” (Unified Facilities Criteria Program 2008).
This refers to the fact that the pressure wave washes around these slender tall
members, and consequently the entire duration of the pressure wave does not act
upon them. Figure 10.9 shows an exposed column support on the corner of a
high rise office building.
For columns subject to a vehicle weapon threat close by in the public street
in an urban area, buckling and shear are the primary effects to be considered in
EXTERIOR WALLS
287
Figure 10.9 Example of column freely exposed to public access. The column supports
the corner of a high-rise building.
analysis. This is a common scenario considered for office buildings in urban areas. Because exposed columns are often in public areas and are primary support
members for buildings, the concern is that progressive collapse may be initiated.
A very large weapon close to a column can cause shattering of the concrete
due to multiple tensile reflections within the concrete section, destroying its integrity. Circular columns shed load more rapidly than rectangular columns and
can be beneficial.
Buckling is a concern if lateral support may be lost when a floor system that
provides lateral support is damaged. This is a situation that arises frequently for
office buildings that are close to public streets. In this case, exterior columns
should be capable of spanning two or more stories without buckling.
Insetting the first line of columns a few feet into the building interior is a
nonconservative approach to protecting the columns. Even if the columns are
not publicly accessible, they are still vulnerable to air-blast effects, and their
response still needs to be addressed.
Protect columns by using closely spaced closed ties or spiral reinforcing, to
improve confinement and shear capacity. This also will improve the performance
of lap splices in the event of loss of concrete cover, and greatly enhance column
ductility. The potential cost-to-benefit ratio for providing closely spaced closed
ties in exterior concrete columns is among the lowest and should be considered
seriously. Closed ties are to be used in columns and spandrels for the entire span
and across connections with a maximum spacing of d/2 and a minimum bend
angle of 135◦ . Some other recommendations for reinforced concrete are given in
wall subsection 10.4.1.
For exposed steel columns, splices should be placed as far above grade level
as practical, and base plates should be recessed below the ground level.
288
BUILDING ENVELOPE AND GLAZING
For a contact or close-in package weapon, column breach is a major consideration. To mitigate this threat, some suggestions include:
r Do not use exposed columns that are fully or partially accessible from the
building exterior. Arcade columns should be avoided (see Figure 10.10).
r Use an architectural covering that is at least 6 inches from the structural
member. This will make it considerably more difficult to place a weapon
directly against the structure. Because explosive pressures decay so rapidly,
every inch of distance will help to protect the column.
r Use steel plates to surround the base of concrete columns where it is accessible. Plates need to extend several feet above the accessible location.
r Encase steel columns with concrete and add a plate, 3/8 –1/2 thick around
the perimeter for about 5 feet above the base, if necessary.
Load-bearing ductile reinforced concrete wall construction without columns
can provide a considerable level of protection if adequate reinforcement is provided to achieve ductile behavior. This may be an appropriate solution for the
parts of the building that are closest to the secured perimeter line (at scaled ranges
less than 3).
Spandrel beams of limited depth generally do well when subject to air blast.
In general, edge beams are very strongly encouraged at the perimeter of concrete
slab construction, to afford frame action for redistribution of vertical loads and
to enhance the shear connection of floors to columns. Confinement of concrete
spandrels using closely spaced closed ties, such as those used for columns, is
recommended. Transfer trusses and transfer girders are to be avoided because
they increase susceptibility to progressive collapse.
Figure 10.10 Example of public walkway beneath second floor of high-rise building
ROOF SYSTEMS
289
10.5 ROOF SYSTEMS
The primary loading on the roof is the downward air-blast pressure. The exterior bay roof system on the side(s) facing an exterior threat is the most critical.
Roof systems that are low and therefore closer to the explosion will be subject to
higher pressures than high roof systems. The air-blast pressures on the interior
bays are less intense and may require less hardening. Secondary loads include
upward pressure due to the air blast penetrating through openings, and upward
suction during the negative loading phase. The upward internal pressures may
have an increased duration due to multiple reflections of the internal air-blast
wave. It is conservative and recommended to consider the downward and upward loads separately.
Because roof systems are not exposed to the reflected wave from the ground
level, they are subject to lower pressures than the walls facing the explosion.
Because the angle between a sloped roof and the advancing shock sometimes
is less acute than the angle between a flat roof and the advancing shock front,
sloping roofs may be subject to somewhat higher pressures. They are usually of
lighter construction than flat roof systems, making them particularly vulnerable
to air-blast effects unless designed to respond in a ductile manner. For this reason,
sloped systems sometimes require more robust designs.
10.5.1 Concrete
The preferred system is to use cast-in-place ductile reinforced concrete with
beams in two directions. If this system is used, beams should have continuous,
symmetrical top and bottom reinforcement with tension lap splices. Shear ties
should develop the bending capacity of the beams and be closely placed along
the entire span. All ties are to have a 135-degree bend minimum. Two-way slabs
are preferred.
Precast and pre/post-tensioned systems, including hollow plank, are generally
viewed as less desirable due to the lack of ductility. If they are used, a system
that has continuous bond with the concrete is preferred, with anchors that are
designed to be protected from direct air-blast effects. Also, additional mild reinforcement at top and bottom is recommended to ensure a ductile response.
Connections need to be designed to resist both the direct and uplift forces.
Concrete flat slab/plate systems are also less desirable because of the potential
of punching shear failure at the columns. Where flat slab/plate systems are employed, they should include features to enhance their punching shear resistance.
Continuous bottom reinforcement should be provided through columns in two
directions, to retain the slab in the event that punching shear failure occurs. Edge
beams should be provided at the building exterior.
10.5.2 Steel
Lightweight systems, such as untopped steel deck, are considered to afford negligible resistance to air blast. These systems are prone to failure due to their low
capacity for downward and uplift pressure.
290
BUILDING ENVELOPE AND GLAZING
10.5.3 Composite
Modest levels of protection are afforded by conventional steel beam construction with a steel deck and concrete fill slab. The performance of this system can
be enhanced by using normal-weight concrete fill, increasing the gauge of deck
and welded wire fabric reinforcement from that required for conventional loads,
and making the connection between the slab and beams using additional shear
connector studs. Tension membrane behavior along the edges should be considered in the design of the connections by using welding or fasteners along the
edges. Since it is anticipated that the slab capacity will exceed that of the supporting beams, beam end connections and supporting columns should be capable
of developing the ultimate flexural capacity of the beams, to avoid brittle failure.
Beam to column connections should be capable of resisting upward as well as
downward forces.
10.5.4 Penthouses/Gardens
Parapets, roof mechanical room enclosures, and tile roof systems are generally
not a primary concern, since they are exterior to the building. Generally these
members are designed to sustain heavy damage but not become flying debris.
Although roofing aggregate may become a flying hazard, in the context of an
explosion event, this hazard is not significant enough to warrant much concern.
Soil can be highly effective in reducing the impact of a major explosive attack.
Bermed walls and buried rooftops have been found to be highly effective for military applications and can be effectively extended to conventional construction.
This type of solution can also be effective in improving the energy efficiency of
the building.
10.6 BELOW GRADE
For buildings that are very close to the secured perimeter, there is the possibility
of the foundations becoming undermined by the cratering effects of an explosion.
However, if this is an issue, then generally, it will be accompanied by heavy damages to the superstructure as well. If the crater is projected to reach the building,
then the most cost-effective option may be to increase the building setback.
Ground shock effects are generally a secondary effect since most of the energy
of a vehicle weapon is transmitted to the air rather than the soil. The weapon
would have to be placed underground to have a significant effect on the structure.
Currently, underground weapons are not considered by the governing federal
criteria for civilian buildings.
There are significant benefits to placing secured areas below grade in terms of
mitigating explosion effects from an exterior weapon. The massiveness and softness of the soil provide a protective layer that significantly reduces the impulse
on the structural systems below grade.
BELOW GRADE
291
For buildings with below-grade portions that are adjacent to the building,
when creating a plaza level at ground level, keep in mind that the roof systems of
these underground areas will need to be designed for the actual air-blast pressure
levels, if these are occupied areas. If these are unsecured areas, such as a garage,
consider letting the roof fail if adequate egress routes are available on other sides
of the building, away from the failed plaza level.
Another consideration for belowground portions of the building is the design
of the perimeter security barriers. The perimeter barriers often require deep foundations, which may interfere with the underground structure.
It is preferable to place underground garages adjacent to the main structure
rather than directly underneath the building, to protect the structure against the
effects of an internal weapon. The effects of an internal weapon are generally
not a major concern for foundation walls. The soil on the other side of the wall
provides a buffer that mitigates the response. One exception to this is a situation
where the foundation is below the water table, where even a localized breach of
the wall may cause extensive collateral damage.
Vaults for transformers placed beneath the ground close to public streets are
of concern. If possible, place these away from public streets. If they are in a
driveway, the vault lid needs to be designed to resist the downward pressure.
In high seismic regions, seismic isolators may be used at the base of a building
(Figure 10.11) . In this case, the response of the building globally should be
checked for the total air-blast loading acting on the side facing the explosion.
Preliminary studies investigating this issue have shown that seismic response
governs.
moat cover
ground floor
ground level
moat
retaining wall
flexible utility connection
basement
base isolator
Figure 10.11 Diagram of seismic base isolation system. The building superstructure is
detached from the foundation and supported by steel plat/rubber bearings.
292
BUILDING ENVELOPE AND GLAZING
10.7 REDUCTION OF BLAST PRESSURES
The placement of the building on the site can have a major impact on its vulnerability. Ideally, the building is placed as far from the property lines as possible.
This applies not only to the sides that are adjacent to streets, but to the sides that
are adjacent to adjoining properties as well, since we cannot be certain about
how access will change for those neighboring properties during the life of the
building. A common-practice example of this is the use of a large plaza area in
front of the building, which often leaves little setback on the sides and rear of the
building. This practice can diminish the vulnerability of the front of the building,
but generally will increase the vulnerability of the other three sides.
The shape of the building can have a contributing effect on the overall damage
to the exterior envelope. Reentrant corners and overhangs are likely to cause
multiple reflections of the air blast, which may amplify the effect of the air blast
(see Figure 10.12).
(a)
(b)
Figure 10.12 Reentrant corner plans: (a) Some plan types. (b) T-shaped building with
two reentrant corners.
REDUCTION OF BLAST PRESSURES
293
In general, convex rather than concave shapes are preferred for the exterior of
the building (i.e., the shock front incidence angle on a convex surface increases
more rapidly with lateral distance from a detonation location than on a planar surface, causing the reflected pressure on the surface of a circular building to decay
more rapidly than on a flat building, See Figure 10.13). Similarly, the air-blast
pressures decay with height, as the angle of incidence becomes more oblique.
The sides of the building not facing toward the explosion do not experience reflected pressure and will typically perform with less damage.
(a)
(b)
Figure 10.13 Building shapes. (a) Convex, and (b) Concave.
294
BUILDING ENVELOPE AND GLAZING
Generally, simple geometries, with minimal ornamentation (which may become flying debris during an explosion), are recommended unless advanced
structural analysis techniques are used. If ornamentation is used, it is recommended that it consist of a lightweight material, such as timber or plastic, which
is less likely than brick, stone, or metal to become lethal projectiles.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
and Commentary (ACI 318-08), Chapter 21. Farmington Hills, MI: American Concrete
Institute.
Amstock, Joseph S. 1997. Handbook of Glass in Construction. New York: McGraw-Hill.
ASTM. 2001. Standard Test Method for Structural Performance of Glass in Exterior
Windows, Curtain Walls, and Doors Under the Influence of Uniform Static Loads by
Destructive Methods (E997-01). West Conshohocken, PA: ASTM International.
. 2003. Standard Practice for Specifying an Equivalent 3-Second Duration Design Load for Blast Resistant Glazing Fabricated with Laminated Glass (F2248-03).
West Conshohocken, PA: ASTM International.
. 2004a. Standard Practices for Determining Load Resistance of Glass in Buildings (E1300-04e1). West Conshohocken, PA: ASTM International.
. 2004b. Standard Test Method for Glazing and Glazing Systems Subject to
Airblast Loads (F1642-04). West Conshohocken, PA: ASTM International.
Computers and Structures, Inc. 1990. SAP2000/NonLinear User’s Manual. Berkeley,
CA: Computers and Structures Inc.
Crawford, John, Shengrui Lan and Brian W. Dunn. 2006. Cable catcher systems for improving blast resistance of glazing facades. In Proceedings of the 19th Military Aspects of Blast and Shock Symposium, Alberta, Canada, October 2006. Burbank, CA:
Karagozian & Case.
Federal Emergency Management Agency. 2003. Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks (FEMA 427). Washington, DC: Risk Management
Series of Publications, FEMA, Department of Homeland Security.
Hyde, David W. 1988. Microcomputer Programs CONWEP and FUNPRO, Applications
of TM 5-855-1, Fundamentals of Protective Design for Conventional Weapons (User’s
Guide). Vicksburg, MS: Army Engineer Waterways Experiment Station, Structures
Lab.
Livermore Software Technology Corporation. 1999. LS-DYNA User’s Manual, Nonlinear Dynamic Analysis of Structures, Version 950. Livermore, CA: Livermore Software
Technology Corporation.
MathSoft. 2001. Mathcad User’s Guide. Cambridge, MA: MathSoft Inc.
MathWorks. 2006. MATLAB: The Language of Technical Computing, Version 7. Natick,
MA: The MathWorks, Inc.
Unified Facilities Criteria Program. 2007. DoD Minimum Antiterrorism Force Protection
Standards (UFC 4-010-01) Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
REDUCTION OF BLAST PRESSURES
295
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Army Engineer Research and Development Center (ERDC). 2001. Window Fragment Hazard Level Analysis, the HAZL Model. Vicksburg, MS: U.S. Army Engineer
Research and Development Center.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
11
Protection of Spaces
MeeLing Moy and Andrew Hart
Security can be integrated at the earliest stage of planning and continued throughout the design and construction phases for building structures requiring blast resistance, or other specialized security measures. Building occupant emergency
plans may also be included in planning such that the design created can supplement the objectives of emergency operations. The goal is to increase the chances
that emergency systems will remain operational during an explosion. The continued operation of particular building elements during emergencies is essential
for building occupant life safety and evacuation. This chapter will focus on the
following building components: areas isolating interior threats, stairwell enclosures, hardened plenums, and safe havens.
11.1 AREAS ISOLATING INTERIOR THREATS
Interior explosive threats may be carried or delivered into accessible areas within
a building. These areas include, but are not limited to, loading docks, mailrooms,
and lobbies. The magnitude of interior threats and the specific areas isolating
them within a building are identified by a risk assessment. The Federal Emergency Management Agency (FEMA) defines a small explosive to be approximately 5 to 10 pounds of TNT equivalent, a large explosive to be approximately
50 to 100 pounds of TNT equivalent, and a mail bomb to usually be less than 10
pounds of TNT equivalent.
Protection of critical building components and systems within the identified
interior high-risk areas may be achieved by installing electronic surveillance
cameras, by implementing controlled access with entry and/or exit checkpoint
screening, and by structural hardening.
Walls and slabs enclosing specific interior high-risk areas should be designed
or hardened for the identified interior threats as determined by a risk assessment.
FEMA recommends that blast effects should be vented through the walls isolating interior threats. For example, these walls may consist of blowout panels
designed to provide security from blast loads applied on the outside due to an
exterior threat, but to fail due to an interior threat, thus venting the blast loads
297
298
PROTECTION OF SPACES
outside. FEMA suggests that slabs isolating interior threats should be designed
or hardened to consider downward and upward blast loads.
11.2 STAIRWELL ENCLOSURES
Exit and emergency egress stairwells are considered to be spaces that are accessible to the general public. They are designed for optimal performance with sufficient distribution and redundancy to meet current building code requirements,
and building occupant needs. Stairwell planning and design are a fundamental
part of a building occupant emergency plan. FEMA recommends that stairwells
should maintain positive pressure for safe evacuation of building occupants, and
for access by firefighters and rescue responders during an emergency. The installation of special filtering systems to provide a clean source of air may be required,
to maintain positive pressure and minimize smoke and hazardous gases in stairwells. The National Institute of Standards and Technology (NIST) recommends
that stairwell design capacities should also take into account the counterflow of
rescue personnel during an emergency. Stairwell lighting should remain functional during an emergency, and stairwell exit signs should provide a clear and
continuous route to the outside or to another area that is safe. In addition to the
building code requirements, stairwells should be located in less vulnerable areas,
and far from areas where blast events may occur in the building. Thus, stairwells
should not discharge into high-risk areas.
Exit and emergency egress stairwell enclosures, which include walls and
slabs, should be designed or hardened for the identified threats specified in a
risk assessment. FEMA recommends that stairwell walls should be anchored to
the floor slabs at each end, in order to adequately transfer the loads from a blast
to the lateral resisting structural system of the building. FEMA suggests that
stairwell slabs should take into account the uplift pressures from a blast by using an upward load equal to the dead load plus half the live load for the floor
system. FEMA also recommends that doors to the stairwell enclosures should
consist of steel, and be adequately anchored to the stairwell walls. FEMA also
suggests that windows in the stairwell enclosures consist of laminated glass with
a 60-mil polyvinyl butyral (PVB) interlayer. The laminated glass should be adhered within the window frames and mullions using a 1/2-inch bead of structural
silicone. The window frames and mullions should be anchored to the stairwell
walls in order to develop the full capacity of the window glazing.
11.3 HARDENED PLENUMS
Mechanical equipment and systems are considered to be primary nonstructural
elements that are necessary for the life-safety procedures of a building. Their
configurations and capacities are designed with adequate redundancy to meet
current building code requirements and building occupant needs. They are
SAFE HAVENS
299
identified as critical assets that should continue main operations during an emergency, to ensure the safe evacuation of building occupants. FEMA recommends
that public access to critical mechanical equipment and systems should be
restricted, in order to reduce the potential damage from a blast event. In addition
to the building code requirements, primary and redundant mechanical equipment
and systems should be situated in separate locations and far from high-risk areas,
to reduce the possible hazard of both being damaged by a single blast event.
Structural elements supporting the plenum that surrounds the mechanical
equipment should be designed or hardened to resist reflected blast pressures
and impulses, such that debris and fragments are not propelled into the protected spaces. Mechanical vents and exposed plenums are designed or hardened
to the same level of protection as required for exterior walls. Designs should take
into account the potential infill of blast overpressures through mechanical openings. Structural elements and mechanical equipment mountings and attachments
should be designed to resist these internal pressures.
11.4 SAFE HAVENS
A safe haven can be defined as a secure area designed to protect occupants from
various hazards. Such shelters may be located outside or within a structure. Safe
havens are fully enclosed structures designed to resist the effects of blast loads,
and the impact of fragments and debris due to an explosive event. This section
will focus on safe havens as documented by FEMA.
11.4.1 FEMA Documents
The FEMA document that deals with the protection of safe havens is primarily FEMA 453, Safe Rooms and Shelters, Protecting People Against Terrorist
Attacks, dated May 2006. Within the document two types of shelters are defined:
r Standalone shelter
r Internal shelter
The difference between the two types of shelters is that a standalone shelter
is a defined as a building that is not attached to or within any other building; it
is considered as a separate building. This separate building is to be constructed
to withstand a range of natural or man-made hazards and may be situated away
from any debris or potential fragment hazards. The building is to be structurally
separated from any other building, and thus not be affected by any weakening
from an adjacent building collapsing. Since this is a separate building, it does
not need to be integrated into the building design, but it is costly compared with
an internal shelter.
An internal shelter is a specially designed room within or attached to a building. The room is designed to be structurally independent of the building and is
300
PROTECTION OF SPACES
able to withstand a range of natural or man-made hazards. Since the shelter is
partially shielded by the building, it may not get the full blast pressure. The location of the room within the building has to be easily accessible by all building
occupants, and when designing for blast mitigation, no resistance from any of
the surrounding areas should be included. Having it within the building is less
costly than building a standalone shelter, as it can be part of a renovation or a
later addition to the building design.
11.4.2 Multi-Hazard Threats
FEMA 453 does not quantify a specific threat such as a terrorist threat from a
vehicle bomb; rather, the document gives general guidelines on different types of
building construction and reasonable mitigating measures for providing a secure
shelter.
Prior Warning In some instances, prior to an event, a warning is given. All
types of warnings given about any type of threat should be treated in a similar manner. The standard procedure for a prior warning is laid out in FEMA
453: Once the warning has been received, the contacted person calmly asks for
the location of the weapon (e.g., the building where it is housed), the type of
weapon, and any other pertinent information that he or she can obtain. Once
obtained, the contacted person must immediately call his/her local enforcement
agencies—such as local and federal law agencies—and emergency services, and
pass on all information from the perpetrator. All warnings should be treated as
reliable, regardless of their apparent validity.
Sequential Events after/during an Event Every type of shelter, regardless of
its size and number of occupants, should have an Emergency Operations Center,
or EOC. This center or group of people serves as an information management
center where decisions are made, such as those made during the attacks on the
twin towers on September 11, 2001. Each EOC should be equipped with communication equipment, reference materials such as medical books, log books to
note down what happened and when, and any other tools necessary to respond
quickly and appropriately to an emergency, such as a defibrillator and a first aid
kit. Within this center, the Emergency Management Group (EMG) makes decisions based upon the information provided by the Incident Commander (IC) and
other personnel. The relationship between the EMG and the emergency services
or Emergency Operations Group (EOG) is shown in Figure 11.1. Note that both
the EMG and the EOG have directors, each in constant contact with each other
throughout the duration of the event, coordinating the responses to the event as it
unfolds.
The use of an IC is vital, as this person is responsible for the front-line
management of the response to the incident, tactical planning and execution,
determining if help is required, and relaying requests for assistance, if required,
SAFE HAVENS
Emergency Operations Group
(EOG)
301
Emergency Management Group
(EMG)
Facility Manager
Emergency Director
Emergency Director
Safety Officer
Operations Officer
Affected Area
Manager/Supervisor
Coordinator Team
consisting of security,
environment,
maintenance, human
resources, planning,
logistics, and public
relations
Emergency Medical
Team
Emergency Services Team
(Fire/HazMat/Police)
Figure 11.1 Relationship between the EMG and the EOG
through the EOC. It is recommended that the IC be a member of management
with the ability to make critical time decisions and to act on them.
Subsequent Fires Because of the lack of fuel, fires from an explosion event
do not usually occur unless the weapon is an incendiary-type device or the event
has set off a chain of events that have given rise to a fire. In the event of an
incendiary bomb, whose sole purpose is to cause a fire, a fuel source close by is
required to sustain the fire. In the case of a chain reaction, the fire usually results
not directly from the explosion itself but from a different source of ignition, such
as a hot fragment igniting a burst gas pipe. Again, these fires, unless they have an
external fuel supply, are usually small and can be easily extinguished by either
cutting off the fuel supply or suffocating the flame.
11.4.3 Design Requirements for Protective Shelters
As in any design, protective shelters need to follow the basic construction codes:
American Concrete Institute (ACI) 318 for concrete (American Concrete Institute 2008), American Institute for Steel Construction (AISC) 360-05 for steel
construction (American Institute for Steel Construction 2005).
It is advisable, before beginning construction of a protective shelter, that a preliminary threat analysis be performed, by the designer, potential shelter owner,
or other responsible party, utilizing the Building Vulnerability Assessment
Checklist included in FEMA 426, Reference Manual to Mitigate Potential
302
PROTECTION OF SPACES
Terrorist Attacks Against Buildings (Federal Emergency Management Agency
2003).
When the preliminary threat analysis has been performed, an independent design professional should perform a more thorough assessment of the shelter to
either confirm the preliminary assessment or modify it accordingly if any deficiencies are found within the design/assessment.
If an existing building is to be used as a shelter, an assessment utilizing the
FEMA 426 Checklist will help to determine the building vulnerabilities and thus
aid in its modifications. It should be noted that FEMA 426 provides guidance to
the building construction industry in how to reduce physical damage to buildings.
The purpose of the manual is to show various approaches that can be taken to
minimize the effects of terrorist threats.
Resistance to Progressive Collapse Progressive collapse occurs when a localized failure occurs and the adjoining members are overloaded. This causes
failure of the members, and a cascade or “pancaking” failure effect occurs. This
failure may result in a partial or total failure of the structure. The initial localized failure could be the result of a small parcel bomb in direct contact with a
critical structural element, or the source could be a vehicle-borne explosive device located a relatively short distance away. A larger explosive device located a
longer distance away is not likely to cause a single member to fail, but will probably affect a number of structural elements instead, which is not deemed failure
by progressive collapse. In addition, according to FEMA 453, Safe Rooms And
Shelters: Protecting People Against Terrorist Attacks (Federal Emergency Management Agency 2006), progressive collapse is not an issue for buildings with
three stories or fewer.
The use of transfer girders and nonductile, nonredundant construction is prohibited in the building construction, as it may give rise to structural systems
that are not tolerant of localized damage. The columns that support transfer
girders—and the transfer girders themselves—may be critical to the stability of
a large floor area. For interior shelters, the building surrounding the shelter must
be sufficiently hardened so that progressive collapse is not an issue. For these
buildings, it is required that the structural integrity of the building must still be
intact if a member fails.
Some of the characteristics that a building containing a shelter must have to
resist progressive collapse are:
r Mass: The larger the inertial resistance, the more resilient it is to failure.
r Shear Capacity: Connections and members need to be designed to prevent
failure.
r Capacity for Load Reversals: Some members may undergo several large
deformations before damping down. These load reversals can be in the form
of uplift pressures, which defy the normal convention of gravity load design.
Structural elements subject to blast effects, therefore, should be designed
for load reversal. For design purposes, it is usualto provide equal areas of
SAFE HAVENS
303
reinforcing steel on both sides of reinforced concrete members, to provide
the necessary resistance. The connections also need to be designed for load
reversals.
r Redundancy: The use of an alternative load path in the vertical load-carrying
system. This allows the gravitational loads to redistribute in the event of a
structural member failure.
r Ties: The use of an integrated system of ties can serve to redistribute the
loads. The ties need to be located perpendicular to the principal line of
structural framing.
r Ductility: The ability of a structural member or connection to have a large
deformation while still maintaining strength. Allowing the member to inelastically deform (e.g., plastically deform) causes large amounts of blast
energy to dissipate; however, special detailing is required.
Using the above guidelines for designing buildings with shelters, or shelters themselves, robustness is greatly improved and progressive collapse may
be avoided. For existing structures that are being modified to accommodate a
shelter, it may be possible to allow only one floor adjacent to the failed column
to fail, but if the structural members are retrofitted to develop catenary behavior,
the adjoining bays must be upgraded to resist the lateral forces. This may result
in more extensive retrofitting than is feasible or desirable. When such cases occur, isolate this region and risk the collapse of the adjoining bays by upgrading
the vulnerable localized columns. When upgrading these columns, a balanced
design approach must be obtained. It should be noted that upgrading an existing
building for an alternative load path system is potentially counterproductive.
For buildings that can provide a continuous load path that will support all
vertical and lateral loads, all structural components and fasteners used in the
connection system must be able to develop the full capacity of the member. In
order for this to take place, the capacity of each component must be balanced
with the capacity of the other members and their connections. As all applied
loads must be transferred eventually to the foundation/ground, the load path must
be continuous from the top structural member to the ground.
Location within the Building One of the most important factors when dealing
with a shelter is its location, including determining the best location for saving
lives. The location of the shelter will depend on how many occupants the shelter
has to hold. In addition to the effects of blast, the location should also consider
effects of chemical, biological, and radiological threats. The shelter should be
located such that all persons who are to take refuge can reach it within a minimal
travel time from where they are located. A way to ease travel time is to clearly
mark the route to the shelter.
Standalone shelters should be located as far as deemed possible from the surrounding buildings, to avoid both progressive collapse and/or impacts from the
collapse of nearby buildings. Internal shelters require that the main building be
designed for progressive collapse.
304
PROTECTION OF SPACES
Access to Egress Shelter locations should be such that all designated persons may take refuge within the shelter with a minimum of travel time. Travel
routes to the shelter should be clearly marked, allowing easy access. Exit routes
from the shelter should direct the persons away from the threat, and hazards
signs should be located using Crime Prevention through Environmental Design
(CPTED) principles. These principles make use of natural access control, natural
surveillance, and territoriality, which can also help to deter the aggressor.
Access to the building must be as “natural” as possible, in that the siting of
the entrances, exits, fences, lighting, and landscaping tends to guide people who
are entering or exiting the building. This is called “natural access control,” and
when it is coupled with limited access, the opportunity for a terrorist attack is
greatly reduced. The placement of external features such as bushes, trees, and
fountains must also be considered so that visibility is maximized. This is termed
“natural surveillance,” and it can reduce the opportunity for terrorists to attack,
by revealing their intentions; in addition, it gives a feeling of safety, as the public
can be easily seen. Finally, the use of fences, bollards, signage, and the like,
termed “territoriality,” clearly defines the property lines between what is public
and private. This can be a gateway into a community or neighborhood.
Fire Rating There are only two fire ratings, H and A (Steel Construction Institute 1990). The H fire rating is solely for hydrocarbon fires such as those caused
by a petroleum product. The A rating is for normal fires such as those that have
wood or paper as their source of fuel. A number usually follows the letter, such
as 60 or 120, and the value represents the time that structural integrity can be
maintained during a fire—e.g., how long a member or area can be kept below a
designated temperature, usually 40◦ C, before the temperature rises, causing that
member to plasticize and fail. In H-rated fires, the maximum temperature is designated as 1100◦ C within 10 minutes, and for A, the maximum temperature is
950◦ C, over 60 minutes (Steel Construction Institute 1990). See Figure 11.2. It
is usual for commercial buildings to be designated as A-rated, as the chances of
a hydrocarbon fire within one of these buildings is remote.
Currently, the fire protection of buildings, based on FEMA, is based on a
four-level hierarchy that comprises alarm and detection, suppression, compartmentation, and passive protection.
r Fire alarms and detectors are usually used to activate the system and notify
the building occupants and emergency services through verbal and visible
systems.
r Suppression systems, usually in the form of water sprinkler systems, are
primarily designed to control small to medium fires and to prevent the fire
spreading beyond the water zone, which is typically 1,500 square feet.
r Use of compartmentation will mitigate the spread of more severe fires.
These typically require fire-rated partitions between the rooms and can
cover an area of 12,000 square feet.
REFERENCES
305
Temperature Curves
1200
Temperature (C)
1000
800
A Rated
H Rated
600
400
200
0
0
5
10 15 20 25 30 35 40 45
Time (min)
50 55
60
Figure 11.2 Temperature profile of H- and A- Rated Fires
r Passive protection is the use of a protective covering on a steel structure.
This protection system will ensure that the temperature of the steel system
does not reach an undesirable level such that the steel becomes plasticized
and cannot withstand the vertical loading applied to it without buckling.
There are numerous different types of ratings for these types of products.
The purpose of this system is to prevent structural failure of the building
before the rated time of the product has elapsed, by which time the building
can be safely evacuated.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
and Commentary (ACI 318). Farmington Hills, MI: American Concrete Institute.
American Institute of Steel Construction. 2005. Specification for Structural Steel Construction (AISC 360-05). Chicago, IL: American Institute of Steel Construction.
Federal Emergency Management Agency. 2003. Reference Manual to Mitigate Potential
Terrorist Attacks against Buildings (FEMA 426). Washington, DC: Federal Emergency
Management Agency, Department of Homeland Security.
. 2006. Safe Rooms and Shelters: Protecting People Against Terrorist Attacks
(FEMA 453). Washington, DC: Federal Emergency Management Agency, Department
of Homeland Security.
Steel Construction Institute. 1990. Interim Guidance Notes for the Design and Protection
of Topside Structures against Explosion and Fire (SCI P-112). Silwood Park, Ascot,
Berkshire, UK: Steel Construction Institute.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
12
Defended Perimeter
Joseph L. Smith and Charles C. Ellison
The four main functions of any comprehensive physical security program are
to deter, detect, delay, and respond to a threat. A defended perimeter serves
an important role in achieving these aims. A well-planned perimeter can deter
an aggressor by increasing the perceived difficulty of attack, can support the
implementation and function of detection sensors, and can delay an aggressor,
providing the security force time to respond to an attack.
12.1 GOALS
The perimeter barrier system will generally be the outermost zone in a tiered
system of protections that includes the building facade and internal screening
zones such as visitor lobbies, mailrooms, and loading docks. These zones, as
shown in Figure 12.1, can be viewed as concentric layers of protection.
In terms of the primary topic of this book, the goal for each tier should be to
reduce the size of the likely explosive threat to a level that can be mitigated by
reasonable levels of blast hardening. As the first line of defense, the perimeter is
the foundation on which all other blast-mitigation options are selected. As such,
the defended perimeter must address the largest threats to the facility. Failure
of the perimeter to perform as expected can have severe consequences for the
adequacy of subsequent blast-mitigation methods.
Heavy barriers are not the only option when creating a defended perimeter.
Perimeter protection may include effective site planning and landscape design
that, if deployed properly, can substantially decrease the risk to occupants and
mitigate damage to a facility. The effective use of landscaping and terrain can
also minimize the design requirements and cost of other man-made barriers employed to stop vehicles or other threats.
12.2 STANDOFF
The standoff distance is generally considered to be the range between the center
of the explosive device and the nearest structural element requiring protection
307
308
DEFENDED PERIMETER
Figure 12.1 Zones of Protection
(i.e., the target). The term “setback” is used when discussing the distance between a facility and the defended site perimeter. These terms are often used interchangeably; however, standoff and setback do not necessarily have to be equal.
Since blast loads decay rapidly with distance, setback can be an effective
means of protecting a facility from an explosive attack. The first few feet provide
the most benefit. Figure 12.2 illustrates peak pressure versus standoff curves for
a range of explosive devices [Based on Equations in UFC 3-340-02, Structures to
Resist the Effects of Accidental Explosions (Unified Facilities Criteria Program
2005b)].
As shown in Figure 12.2, at very small standoff distances (i.e., less than 5 ft)
just an extra foot of additional standoff can reduce the blast load on a structural
member by an order of magnitude (i.e., a factor of 10) or more. It may only take
a few pounds of explosive in direct contact with a structural column or wall to
cause devastating structural collapse. However, that same column or wall may
resist several hundred pounds of explosives if only a few feet of standoff are
provided. The benefits of a few additional feet of standoff can also be significant
at much greater distances. For example, the pressure loads generated by a 500-lb
TNT bomb at 112 ft are over 20% less than the loads generated at 100 ft.
Even though the benefit per foot of standoff decreases with distance, providing standoff is often more cost-effective than structural hardening for significant
distances from the facility. To illustrate this point, the approximate relationship
STANDOFF
309
10,000
5 lb TNT
50 lb TNT
500 lb TNT
5,000 lb TNT
50,000 lb TNT
Pressure (psi)
1,000
100
10
1
0.5
0
100
200
300
400 500 600
Standoff (ft)
700
800
900 1,000
Figure 12.2 Blast Pressure Loading versus Standoff
between defended standoff and protection offered by conventional construction
is shown in Figure 12.3 (Smith et al. 2005).
Figure 12.3 illustrates the level of protection offered by a wide range of conventionally constructed buildings for a given setback. The left portions of the
bars indicate that no significant protection from blast effects is readily attainable
at these distances with conventional construction. The next area is an indication of a low level of protection. At these distances, conventionally constructed
buildings will typically sustain moderate to heavy damage. Occupants in exposed structures may suffer temporary hearing loss and injury from the force
of the blast wave and building debris fragmentation. Other assets may receive
damage from these effects. The third area is an indication of a medium level
of protection. At these distances, conventionally constructed buildings will generally sustain light to moderate damage. Occupants of exposed structures may
suffer minor injuries from secondary effects such as building debris. The rightmost area indicates a high level of protection. At these distances, conventionally
constructed buildings will generally sustain only minor damage.
12.2.1 Balancing Hardening with Standoff
A common question that designers face is “Where do I put my barriers?” Typical security personnel will declare that barriers must be as far away from the
protected asset as possible. Typical architects will respond that barriers intrude
310
DEFENDED PERIMETER
Medium-Level
Protection
No Protection
Low-Level
Protection
High-Level
Protection
Explosive Weight (lbs.)
1000
500
220
50
0
200
400
600
Standoff Distance (Feet)
Figure 12.3 Effects of Standoff Distance for a Typical Conventionally Constructed
Facility (explosive weight is in lbs of TNT equivalent)
on the appearance and function of the project and must be minimized. Typical
owners will say they want to be safe but, at the same time, not inconvenienced.
Typical community planning boards will want free, unrestrained access to the
site for community use. In most practical cases, especially for government facilities, there may be design criteria that prescribe a specific minimum standoff for
the defined threats. In other cases, there may be no such clear requirement. In the
end, politics or other nontechnical factors may also play a deciding role in the
selection and placement of barrier systems.
So what’s a designer to do? It is important to consider all aspects of the
design in determining where barriers are placed, the type of barriers to be used,
and the protection level that such barriers will provide. The site features may
play a role in determining the placement of the protected asset or building
within a protected site. From a protection engineer’s perspective, balancing
setback requirements with structural hardening and hazard-mitigating features
should be of primary concern. Designing to resist explosive effects increases the
STANDOFF
311
requirements of structural components; thicker walls, additional reinforcement,
and blast-resistant glazing and frames may be needed. An increase in explosive
weight or a decrease in standoff generally increases structural requirements. For
an efficient design, one must balance the effect of available standoff with the
effect of incorporating blast-resistant design and/or hazard-mitigation measures.
The types of perimeter protection should also be balanced with the types of
attacks considered in the risk assessment. A crash-rated barrier line is not needed
if the attackers are unwilling to risk drawing attention to themselves. However,
heavy barriers may be the only option to stop a truly determined attacker. Potential attacks can generally be reduced to three types:
r A true covert attack in which the attacker desires to completely escape notice and repercussion. Attackers of this type will avoid anything that draws
attention to themselves, such as high speeds and illegal parking. Eliminating public parking near the building and presenting a notable guard presence
will deter this type of attacker and provide standoff.
r An attack in which the aggressor is dedicated to the attack, but intends to
escape. This attacker will break traffic laws, but wants to get out of the
area before a police response can be mounted. Designated parking or even
illegal parking will not deter this attacker since he or she can flee the area
before a response is mounted. However, heavy-rated barrier systems are not
necessarily required, since this type of attacker will be deterred by the risk
of a potentially fatal crash or a significant delay in escaping.
r A suicide attack by an attacker who is willing to be captured or killed in
the attempt. If vehicle barriers are not in place, this attacker will drive onto
the site and through the front door of the building. Stopping this threat will
require significant and consistent hardening of the perimeter against vehicular attacks. Developing appropriate limits on achievable barrier penetration
is crucial. Sally ports and other redundant barriers may be necessary at all
vehicular entrances.
12.2.2 Balancing Costs
The effects of standoff provided by barrier systems on various structural and
nonstructural components are conceptually illustrated in Figure 12.4 (Smith
et al. 2005). This figure generally illustrates, at no specific scale, the typical
relationships between standoff and the cost of protection. A number of the
various components of incremental security cost are shown, including structural
and nonstructural component contributors. The relative magnitude and scale of
these relationships will vary from project to project.
For example, the cost associated with hardening the mailroom, loading dock,
and lobby to meet the design requirements may not vary with the available standoff to a vehicle-delivered bomb along the site perimeter. The cost associated with
progressive collapse considerations may also be constant with standoff, since it
312
DEFENDED PERIMETER
Incremental Cost of Protection $
is often treated as threat-independent by security design criteria. However, there
may be a point at shorter standoffs where the structural framing design is further impacted by the blast loading on the frame (above the progressive collapse
requirements), resulting in larger framing members and additional cost. Some
criteria may place limits on the design blast pressure and impulse for certain
secondary building components such as doors and windows, resulting in the relatively flat curves seen in the figure. However, some criteria do not use such restrictions and are likely to experience a significant cost increase at low standoffs.
The sum of the varying costs of hardening for the various components results in
the “cost of hardening” curve indicated on Figure 12.4.
One cost component that increases with increasing standoff is that for site
area and perimeter protection. For example, to provide increased standoff, the
distance to the defended perimeter must increase, thereby increasing the area of
Total Protection Cost
(hardening + land + perimeter)
Not To Scale
cost of land +
perimeter protection
cost of hardening
frame
progressive
collapse
windows & walls
other, mailroom, loading dock, lobby
20
50
limit
STANDOFF (ft)
RISK
High to
Catastrophic
Moderate
Moderate to Low
High to Moderate
Figure 12.4 Impact of Standoff Distance on Total Protection Costs (Compliments of
Applied Research Associates, Inc.)
STANDOFF
313
the site and the length of the perimeter that must be protected. Adding the cost
of hardening and the cost of land and perimeter protection results in the curve
indicated as “Total Protection Cost.” In the particular example shown in Figure
12.4, a standoff distance in the range of 30 to 50 feet would provide the desired
protection at minimum total cost. A project-specific analysis that examines these
variables can help in determining the optimal standoff for the proposed site.
12.2.3 Site Planning
Good site planning is the first step in ensuring that the desired standoff distances
are achieved at a given site. An effective site plan will minimize or eliminate
places where explosive devices can be concealed, will ensure that all activity
around the facility is subject to casual observation by the occupants or actively
monitored by a security force, and will discourage or prevent pedestrian or vehicular traffic in vulnerable areas.
There are an almost endless number of possibilities when it comes to implementing security features on a new site, and there is a great deal of room for
creativity. As shown in Figure 12.5, numerous types of barriers can be integrated
into the site plan. The implementation of landscaping and architectural features
can also have an important effect on the vulnerability of a facility. The positioning of roadways, parking, and sidewalks relative to the building can discourage
or encourage traffic. These features can also affect the likelihood of detecting
an attack or the time between detecting an attack and responding to it. In the
case of vehicular attack, roadway planning and landscaping features can limit
Figure 12.5 Integrated Site Plan
314
DEFENDED PERIMETER
the impact speed and approach angle of an attacking vehicle, resulting in lighter
barrier requirements. However, it is important that security be considered early
in the design process, when civil, architectural, and structural changes can be
made with minimal impact on ongoing designs in other nonsecurity disciplines.
Controlling the maximum impact velocity of an attacking vehicle is one very
important part of site planning, since the velocity term is squared in the relationship for kinetic energy and has a very large effect on the size and cost of
vehicular barriers. For example, doubling the impact speed increases the impact
energy for a given vehicle impact by a factor of four, thus greatly increasing the
design requirements and potential cost of a barrier system.
Using basic engineering and physics, there are formulas and methods available to estimate the maximum attainable impact speeds that a vehicle may obtain.
Performing such a study early in the planning process can significantly reduce
the costs associated with vehicle barriers.
Attainable Vehicle Speed on a Straight, Level Path The most basic vehicle
speed calculation is based on a straight and level path between the starting point
and the vehicle barrier. The final speed for such an approach can be calculated
using the following basic equation (Unified Facilities Criteria Program 2005b):
v2f = v20 + 2as
where: vf = final vehicle velocity
v0 = initial vehicle velocity
a = constant rate of acceleration
s = distance traveled
Initial vehicle velocity and distance traveled are easy concepts to grasp and
develop input values for. However, rate of acceleration can be difficult to determine. First, acceleration is very dependent on the weight and power output of
the vehicle. Average accelerations between zero and 60 mph can range from 25
feet per second squared for a modern sports car to 5.8 ft per second squared for a
2.5-ton commercial truck. Second, acceleration is not constant. Neglecting drag
and rolling friction, a vehicle producing a constant 200 hp at the rear wheels and
weighing 4000 lb will accelerate from 20 mph to 30 mph in 0.61 sec (average
acceleration = 24 ft/secˆ2). However, the same vehicle will require 1.34 seconds
to accelerate from 50 mph to 60 mph (average acceleration = 11 ft/secˆ2). As
can be seen, a 30-mph increase in speed reduced the rate of acceleration by
one half.
These numbers were obtained by solving the following equations through
time:
vi = v(i -1) + a(i -1) (dt)
ai = P/(mvi )
STANDOFF
where:
315
P = power
m = mass
vi = current velocity
v(i-1) = velocity at last time increment
ai = current acceleration
a(i-1) = acceleration at last time increment
dt = time increment
Slopes A downhill slope will decrease the required acceleration distance, and
an uphill slope will increase it. A correction factor can be obtained using the
following equation:
s /s = 1/[1 + (g/a) sin(theta)]
where:
s = acceleration distance needed to attain final speed on a sloped path
s = acceleration distance needed to attain final speed on a horizontal
path
a = acceleration
g = gravitational constant = 32.2 feet per second squared
theta = angle of slope (0 is level, positive is downhill slope)
(Unified Facilities Criteria Program 2005b)
Maintainable Vehicle Speed on a Curved Path Centrifugal force will result in
loss of control of a vehicle driving on a curved path if it exceeds the available
friction force on the tires. By equating centripetal force to friction, the following
equation can be derived for a vehicle traveling on a flat surface:
vs = sqrt( fgr)
where: vs = skid velocity
f = friction coefficient
g = gravitational constant = 32.2 feet per second squared
r = radius of curvature
(Unified Facilities Criteria Program 2005b)
The value of the friction coefficient, f , is highly variable. It depends on the
tire and its condition, the material and condition of the drive path, and any other
factors such as oil or water on the drive surface. The value of f should generally
fall between 0 and 1. For common vehicles under normal conditions, f = 0.6
is usually used. Values greater than 1 are possible for a dedicated race vehicle
on specially prepared tracks, but a value of f = 1 is generally safe when road
conditions are unknown, or a more conservative value is desired.
Curve Banking Banking can increase the achievable speed of a vehicle in a
curve and is commonly used in roadway design. The following equation adds an
316
DEFENDED PERIMETER
additional term to the curved path equation for calculating maintainable vehicle
speed on a banked curve, to cover that possibility (Hickerson 1967):
vs = sqrt[( f + tan(alpha))gr ]
where:
vs = skid velocity
f = friction coefficient
g = gravitational constant = 32.2 feet per second squared
r = radius of curvature
alpha = angle of embankment, positive for outer edge of curve higher
than inner edge
12.3 VEHICLE CONTROL BARRIERS
Since larger explosive weapons are difficult to carry and conceal on an individual, vehicles play a critical role in many explosive attacks, and vehicular control
should be considered when protecting a facility from large explosive threats. A
successful vehicle control system will prevent an unscreened vehicle from reaching a point where it can be used to damage or destroy the facility.
A well-planned site with good detection equipment and a properly trained
guard force can be effective in keeping what is generally referred to as a stationary vehicle-delivered threat away from a facility. That is, a vehicle driven
to the target, parked, and left by an aggressor intent on escape. However, barriers designed to defeat moving vehicular threats will be required to stop a more
determined aggressor willing to be captured or killed in a successful attack.
Barriers designed to defeat a moving vehicular threat (that is a threat actively
driven into the site or facility) must be designed to arrest the energy of the attacking vehicle. For most standards, kinetic energy is used as a basis for establishing
vehicle barrier performance requirements. The gross mass of a vehicle (vehicle
mass plus the mass of explosives or any other cargo) and its maximum attainable
speed at the point of impact define the kinetic energy that must be absorbed by
the barrier. Kinetic energy can be expressed as:
K E = 0.5 mv2
where: KE = kinetic energy
m = gross mass of vehicle
v = velocity of vehicle
12.3.1 Crash Testing
Both the U.S. Department of State (DoS) and the U.S. Department of Defense
(DoD) rate barriers based on full-scale crash tests conducted by independent test
laboratories or government-approved facilities. The testing standard traditionally used for these tests is designated SD-STD-02.1, Specification for Vehicle
VEHICLE CONTROL BARRIERS
317
Crash Test of Perimeter Barriers and Gates (U.S. Department of State 1999).
The DoS test standard rates the anti-ram barriers within three categories: K4,
K8, and K12. The K rating corresponds to the kinetic energy of the impacting
vehicle at the moment of impact. A K12 rating corresponds to impact energy of
1,200,000 ft-lb. These tests are conducted using a flatbed truck with a designated
mass of 15,000 lb. The standard originally provided an additional L rating that
designated the level of penetration achieved by the vehicle. However, in March
2003, with the release of Revision A, acceptable penetration was limited to 1 m,
eliminating the need for an L rating. All previous barriers where the penetration
exceeded 1 m were designated unacceptable by the revised standard, and the L
rating was dropped from use. While significant penetration is not allowed by
the DoS standards, penetration may be permitted in other applications especially
where large setback is available.
ASTM has also developed a test standard, with input from both the U.S. Department of State and the U.S. Department of Defense, that is beginning to grow
in acceptance in the barrier industry. This standard, Standard Test Method for
Vehicle Crash Testing of Perimeter Barriers (ASTM F2656-07) is likely to grow
in use and is gradually being accepted by both the U.S. Department of State and
the U.S. Department of Defense.
12.3.2 Crash Modeling
There is no one-to-one correlation between the kinetic energy and a static design
load. The impact forces vary significantly through time and are a function of the
crushing of the vehicle as well as the strength and stiffness of the barrier. One
method for calculating barrier response is to equate the energy in the vehicle
(KE) with the work done on the barrier and the vehicle:
KE = Wb + Wv
Wb = F(x) · d xb
Wv = F(x) · d xv
where: KE = kinetic energy
Wb = work done on the barrier
Wv = work done on the vehicle
F = force generated between the barrier and the vehicle
xb = displacement of the barrier
xv = displacement of the vehicle
Calculating work on the barrier and the vehicle depends on obtaining good
force and displacement histories. This is difficult to accomplish with simple
models, given the complex interactions that occur between the barrier and the
deforming vehicle during impact. A conservative simplified analysis can be performed using a single-degree-of-freedom (SDOF) model of the barrier, if the
work done on the vehicle is set to zero. However, an SDOF design will generally
318
DEFENDED PERIMETER
Figure 12.6 Comparison of Vehicle Impact Test to Finite Element Analysis (Compliments of Applied Research Associates, Inc.)
be very conservative. Testing has shown that the real force and displacement histories that occur during impact are quite complex, and the energy absorbed by
the vehicle can be significant.
Given the need to make assumptions that ensure conservatism, barrier system designs resulting from simplified analysis can be significantly overdesigned.
Therefore, barrier systems are generally certified through full-scale crash testing
or sophisticated analysis.
Properly performed nonlinear finite element analysis (FEA) techniques have
shown good correlation to test data and can be an effective means of developing
new barrier systems. FEA models can replicate many of the complex interactions
that control the force and displacement histories in actual crashes. Examples of
an actual vehicle impact compared to a successful pretest finite element analysis
are shown in Figures 12.6 and 12.7.
Finite element analysis is a powerful tool, but it should generally be supplemented with crash testing to verify the analysis results. When such testing is not
possible, a qualified firm or consultant who is independent of the manufacturer
should be contracted to perform or review such analysis. FEA is most effective
Figure 12.7 Front View of Impact versus Analysis Result (Compliments of Applied
Research Associates, Inc.)
VEHICLE CONTROL BARRIERS
319
Figure 12.8 Low Wall (Compliments of Applied Research Associates, Inc.)
when the analyst has access to crash test data for similar systems and vehicles in
order to validate the FEA model performance. Therefore, the more the proposed
system differs from known test parameters, the more doubt and scrutiny should
be given to the results. Ideally, the analyst will be able to provide FEA results
that demonstrably replicate tests for similar barrier systems.
12.3.3 Walls
Walls and retaining walls, in particular, are some of the most robust crash barrier systems available. Commercial wall systems are available that have been
designed and tested to stop a 60,000-lb vehicle traveling at 50 mph (KE =
5,000,000 ft-lb). Reinforced concrete is a common construction material for
crash-rated walls. The wall thickness is generally between 12 and 24 inches.
Crash-rated walls have traditionally risen 36 inches above grade, but recent testing and analysis have supported the use of 30-inch-tall barriers for stopping typical cars and some trucks. Figure 12.8 illustrates what can happen when the wall
height is set too low.
12.3.4 Bollards
Bollards are one of the most common vehicle barrier systems in use. Bollards
are usually cylindrical posts spaced close enough to prevent vehicle passage and
are typically constructed of steel pipe anchored in a reinforced concrete foundation. A few bollard systems are designed with foundations less than 1 foot deep,
but most have deeper foundations. Four feet is typical. Bollard spacing is also
frequently set at 4 ft, but bollards can in many cases be effective at spacing up
to 5 ft. Spacings beyond 5 ft are generally avoided, since small cars or trucks are
capable of squeezing between bollards spaced farther than 5 ft apart. A typical
passive bollard design is shown in Figure 12.9.
The bollard shown is capable of a K4 rating. K12-rated bollard designs are
also available.
320
DEFENDED PERIMETER
Figure 12.9 Eight-Inch-Diameter Fixed Bollard System
Active bollards can also be obtained. The most basic can be manually detached from the foundation and carried away (Figure 12.10).
The more advanced systems can be retracted into the foundation, using hydraulic or electric motors. Figure 12.11 illustrates one such retractable hydraulically operated bollard system.
12.3.5 Active Wedge
Wedges are a common type of active barrier and are capable of achieving K12
ratings. Wedges are typically installed to be flush with the road surface when
lowered. When the barrier is raised to stop traffic, steel plates pivot up from the
road surface to form a wedge shape. Wedges are typically operated electrically
or hydraulically. Many variations are available. Figure 12.12 illustrates one common configuration of wedge barrier.
12.3.6 Beam Barriers
Beam barriers generally swing, pivot, or slide out of the path of the vehicle to
permit traffic flow. Beam barriers generally do not achieve impact ratings as high
as those of other active wedge barriers, but they generally cost significantly less
and are easier to operate. The photo in Figure 12.13 shows a beam barrier.
VEHICLE CONTROL BARRIERS
Figure 12.10 Removable Bollard
Figure 12.11 Operable (Pop-Up) Bollards
321
322
DEFENDED PERIMETER
Figure 12.12 Wedge Barrier
Figure 12.13 Beam Barrier
VEHICLE CONTROL BARRIERS
323
12.3.7 Cable-Based Systems
Cable-based systems generally consist of horizontally spanning steel cables anchored to the ground through anchor posts or other means and positioned to
capture an attacking vehicle. When a cable system is impacted by a vehicle, the
intermediate anchor posts respond progressively away from the point of impact
until the vehicle is arrested. Cable-based barrier systems generally allow more
penetration than other more rigid barrier systems. However, they also generally
require lighter or more widely spaced foundations. Another advantage of cablebased systems is their ability to absorb the energy of the impact in a fashion that
limits vehicle damage and injury to the vehicle occupants. The anchored cables
can also be integrated into various fence systems to prevent intrusion of people
as well as vehicles. See Figure 12.14.
Cable barriers are a good choice for large perimeters where a few feet of
penetration is acceptable in exchange for reduced installation costs. Cable-based
systems are most effective for protecting long stretches of perimeter, since they
Figure 12.14 Cable Fence Barrier
324
DEFENDED PERIMETER
Figure 12.15 Concrete Planter
perform best when splice points are widely spaced. Fences with cables are
common at airports throughout the United States. These systems are generally
vulnerable to impacts near corners and ends, where the impact forces cannot be
transferred to adjacent anchor posts through tensile-membrane action. Special
detailing or supplemental barriers may be required near corners to ensure a
consistent level of protection.
Active cable-based systems also exist and typically consist of a cable net that
spans the road to prevent traffic flow. The cables can be relaxed to allow vehicles
to pass over the cable net or tensioned to activate the barrier.
12.3.8 Planter and Surface Barriers
Surface barriers typically consist of large containers filled with soil, water, sand,
or some other readily available material. One of the more common and aesthetically pleasing barriers in this category is concrete planters (see Figure 12.15).
True surface barriers (ones that do not include a foundation) initially absorb
the energy of impact through momentum transfer and must have significant
mass to perform well. Lighter surface barrier systems may allow significant
penetration before stopping a vehicle. Figure 12.16 shows the aftermath of a
real-world collision with a relatively lightweight planter system. In this case, the
planter was driven nearly to the building before coming to a full stop. Ground or
concrete anchors are available for such systems and will significantly reduce the
penetration distance.
12.3.9 Berms, Ditches, and Other Landscaping Features
Berms, ditches, and hillside cuts, as well as trees, boulders, and other large
objects can be effectively used to stop vehicles from penetrating a restricted
PEDESTRIAN CONTROL BARRIERS
325
Figure 12.16 Concrete Planter Penetration
boundary. The authors are not aware of detailed analysis or testing that verifies
the kinetic energy resistance of this category of barrier, but these systems can
provide significant vehicular control. Figure 12.17 diagrams several commonly
used barriers in this category.
Triangular ditches, trapezoidal ditches, and hillside cuts are easy to construct
and are effective against a wide range of vehicle types. The designs presented in
Figure 12.18 are based on typical vehicles with respect to tire size and ground
clearance. Dedicated off-road vehicles may defeat these systems, and careful
consideration should be given to likely threats. It is also important to consider
the speed and trajectory of an approaching vehicle when implementing any of
these systems. For example, it may be possible for a vehicle to jump a ditch
or hillside cut, given an ideal approach. The width of the ditch may need to be
increased where a high-speed approach is possible.
Boulders, trees, and other vegetation can also be used to stop vehicles. One
such application is shown in Figure 12.18.
12.4 PEDESTRIAN CONTROL BARRIERS
Pedestrian control is designed to direct and moderate the flow of people around
or through a facility. Good pedestrian control will deter or prevent visitors or
uncleared personnel from entering areas of the facility vulnerable to attack. It is
important to remember that a properly placed man-portable explosive device in
close contact with a protected asset or structure can have a devastating effect.
Good site planning is the first step in pedestrian control and may be the
only pedestrian control system implemented outside the building envelope for
many facilities. An effective site layout will eliminate places where explosive
326
DEFENDED PERIMETER
Figure 12.17 Excavation Barriers
BLAST WALLS AND BERMS
327
Figure 12.18 Trees and Boulders Used as Barriers
devices can be concealed, will ensure that all activity around the facility is
subject to casual observation by the occupants, and will discourage pedestrian
traffic in vulnerable areas without drawing attention to those areas. Sidewalks,
landscaping, lighting, and signage can all contribute to achieving these goals.
True physical pedestrian barriers are prudent for facilities where there is a
significant risk of pedestrian attacks. Fencing or high walls are the most common
type of pedestrian control barrier for the building perimeter. Fencing and walls
come in many sizes and architectural styles, and each style carries a different
delay factor based on height and structure. However, for the purpose of security,
fencing or walls should generally be a minimum of eight (8) feet high and be
durable enough to impede or withstand attacks from at least simple hand tools
such as bolt cutters. In some cases, fences and walls can be built to also act
as vehicle barriers. An example of this is shown in Figure 12.19. This fence
application includes heavy cables secured to massive, buried anchors that serve
to effectively stop a vehicle-borne threat.
12.5 BLAST WALLS AND BERMS
Obtaining adequate standoff to reduce the blast loads to an acceptable level is not
always possible. One way to reduce loading is the implementation of blast walls.
Properly implemented blast walls and revetments can mitigate the peak air-blast
pressure and impulse from an explosion.
When placed across the path of the air-blast shock wave, the barrier may
reflect a portion of the incident pressure back toward the explosion, or cause
the blast to diffract over the barrier. The diffracted pressure will be significantly
328
DEFENDED PERIMETER
Figure 12.19 Anti-Climb Fence with Integrated Catch Cables
reduced for some distance behind the barrier before the shock reforms to its
original intensity. The area of effectiveness behind the barrier is controlled by
the dimensions of the blast wall or revetment, the size of the explosive device,
and the relative locations of the bomb, the wall, and the building.
Properly designed blast walls can also capture vehicle and bomb fragments
before they reach the building. This can be an important consideration, especially when dealing with military-grade weapons designed to generate hazardous
fragments.
Proper hardening of the wall is an important consideration in implementing a
blast wall. Testing has shown that walls located close to the explosive device can
break up and become dangerous fragments during an explosion. In many cases,
the energy imparted to the facility by wall debris can negate the benefits provided
by the reduction in air-blast loads. Using wall types that are inherently resistant to
generating hazardous fragments is an option. Lightweight containers filled with
water and sand are examples of systems with low secondary fragmentation risks.
The phenomena that occur behind a blast wall and that generate load reductions are very complex. A site-specific analysis is required to properly locate
the wall and calculate the benefits. An improperly implemented blast wall will
provide little benefit, and it is possible for a blast wall to actually increase the
blast loads on a structure. A documented procedure for analyzing the benefits of
blast walls is presented in UFC 4-020-03, Security Engineering, Final Design
(Unified Facilities Criteria Program 2005a).
The primary sources of information about the benefits of blast walls and
berms are experiments conducted on blast walls and a study on revetments.
These studies only covered a limited range of devices and barrier heights, but
the following trends have been observed. The wall must be close to the building
REFERENCES
329
or the explosive charge to achieve useful load reductions. For walls placed
close to the charge, peak pressures are reduced dramatically directly behind
the wall but gradually approach the free-field blast at large distances. There are
some combinations of large distances between the bomb and the wall and short
distances between the wall and the building that may cause an increase in the
reflected impulse due to multiple reflections of the blast wave between the wall
and the structure.
The shape of the wall or barrier and its height have a significant effect on the
load reduction. Walls and revetments with faces oriented perpendicular to the
shock wave (and the ground surface) should be a goal, since this arrangement
provides by far the most benefit. Mounded or sloped barricades (i.e., typical berm
construction) provide little reduction in blast loads.
The current blast wall data are based on situations where the wall height was
selected by taking the cube root of the charge weight in pounds and multiplying
by values between 0.8 ft and 1.3 ft. For example, the cube root of 500 lb is 7.9.
Therefore, 6-ft to 10.5-ft walls fall within the range of the data. Trends in the
data show that the wall is most effective if the bomb is within 1 wall height of
the wall and within approximately 10 wall heights of the building. Ranges from
the wall to the building greater than 20 wall heights generally result in negligible
reductions.
REFERENCES
Hickerson, Thomas F. 1967. Route Location and Design (5th ed.) New York: McGraw
Hill.
Smith, Joseph L. et al. 2005. Protective Design and Security Implementation Guidelines.
Prepared by Applied Research Associates, Inc. for the U.S. General Services Administration.
Unified Facilities Criteria Program. 2005a. Security Engineering, Final Design (UFC
4-020-03). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria
Program.
. 2005b. Selection and Application of Vehicle Barriers (UFC 4-022-02). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
U.S. Department of State. 1999. Specification for Vehicle Crash Test of Perimeter Barriers and Gates (SD-STD-02.1). Washington, DC: U.S. Department of State.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
13
Blast-Resistant Design of
Building Systems
Scott Campbell and James Ruggieri
13.1 BACKGROUND
Public awareness of the need to protect structures against blast effects has risen
sharply since September 11, 2001. A number of public agencies are engaged
in research and development of blast-resistant technology, and its application to
domestic and foreign structures that are potential targets of terrorist activity. The
federal agency responsible for developing technology in this area is the Defense
Special Weapons Agency/Technical Support Working Group (DSWA/TSWG),
which receives funding from Congress through Department of Defense (DoD)
appropriations. This funding is used to support physical testing of candidate
blast-resistant design concepts, computational modeling, and understanding
of phenomenology. A range of government agencies and private consultants
participates in the program, whose results are available to the General Services
Administration (GSA), the DOD, and other U.S. and foreign government
agencies. The technology derived from this and similar previous programs is
currently being implemented for the protection of U.S. Secret Service headquarters and FBI laboratory in the Washington, D.C.. area, the federal courthouse
in Brooklyn, NY, and the passenger terminal at Midway Airport, Chicago, to
name only a few locations. The U.S. State Department plans to request between
$1 and $3 billion to raise security at U.S. embassies to the standards set by the
Department in the 1980s.
The events of 9/11, as well as similar follow-on events, have introduced
new performance requirements for all building systems. New performance
requirements for such structures have introduced a need to revisit and modify
the traditional assumptions governing building design. New systems, including
atmosphere segregation systems (i.e., positive pressure, collective protection,
and hazardous agent monitoring systems, etc.) as well as changes to the
environment, including consideration to blast and shock, introduce constraints
not previously considered in the design development of building mission and
life-safety systems.
331
332
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Although existing commercial consensus standards, such as NFPA 70A (National Fire Protection Association 2005), have historically performed well for
general “nontarget” buildings and facilities, they do not address the special considerations now facing “target” buildings and facilities serving critical infrastructure needs. Therefore, new practical and readily expedient methods need to be
identified that could better fulfill this new environment.
13.2 INTRODUCTION
The detonation of explosives near or in a building affects not only the building
facade and structure, but also the building contents. In seismic events, the
improvement in earthquake-resistant structural design has led to a significant
decrease in structural damage. However, the result is that nonstructural damage
now accounts for a majority of the costs associated with earthquakes. Similarly,
increasing the blast resistance of building structures will reduce the associated
damage, but may increase the relative importance of nonstructural component
damage. Significantly, a building may be structurally undamaged, but rendered
useless due to damage to HVAC, electrical, plumbing, and other systems.
Therefore, a balanced design approach is suggested wherein the building
systems are designed to withstand either the appropriate blast loads or resulting
structural movements.
Very little information is available regarding the design of building systems to
resist the effects of explosions. This lack of design guidance has resulted in the
use of better-developed seismic guidelines, such as the Uniform Building Code
(UBC) (International Code Council 1997) or International Building Code (IBC)
(International Code Council 2006), the use of extremely conservative designs,
or the neglect of the protection of equipment and systems. Traditionally, equipment has been designed for blast by being placed in shelters or internal to the
building and checking acceleration resistance against expected levels. This remains the best practice, although particularly rugged equipment—for example
pumps—may be left exposed with anchorage designed for the applied loads.
Third-party testing by nationally accredited conformity assessment organizations, such as Underwriters Laboratories (UL), is often used to verify certain performance characteristics of a system element. Traditional testing often includes
verification of desired material properties such as strength, ductility, chemistry,
and composition; and random sampling procedures are used to audit performance of the subject components for use, in comparison to design requirements
and/or regulation. Examples of this include equipment operability, controllability and speed, current-carrying capacity, insulation resistance, power required,
power developed, vibration, and many others. In most cases, the requirements invoked for electrical equipment and devices used in building systems only address
electrical shock safety (e.g., electrocution risk) and risk of fire. However, such
criteria presume a benign environment, not subject to risk of blast. The standards
DESIGN CONSIDERATIONS
333
and building code organizations readily admit that such documents only identify
a consensus for a common denominator that speaks to the minimum acceptable
performance or criteria, assuming a safe and relatively comfortable environment.
Such documents do not address blast. Although there is one U.S. national standards development activity presently underway to provide enhanced criteria to
harden building electrical systems to the effects of blast—ASCE/AEI Recommended Hardening Techniques for Control, Communication and Power (C2P)
Systems of Critical Facilities—there is nothing available in the nonmilitary community at this time that can be used for design guidance.
Explosions generate loads on building systems through three basic mechanisms. Direct blast pressures can impact systems exterior to the building from
explosions outside the building envelope and from internal explosions. Indirect
pressures are generated when a blast external to the space where the equipment is located leaks into the room through openings. Finally, ground shock and
air shock generate loads on building systems through motions developed in the
structure. All three types of loads are considered in this chapter. Note also that
fragments can pose a severe threat to building systems and equipment. However,
calculation of fragmentation and the resulting damage to nonstructural components is beyond the scope of this section (see Chapter 8).
The general design approach for building systems is to determine the load
levels, both blast pressures and base accelerations, for the chosen location. The
equipment and anchorage capacities are then checked against the design load.
If the equipment’s allowable load is exceeded, then either it must be moved or
protected. Each of these steps is detailed in subsequent sections of this chapter.
13.3 DESIGN CONSIDERATIONS
The key philosophy for providing continuity of services for mechanical and electrical systems in buildings is SHR: Separation, Hardening, and Redundancy.
Segregation of critical and vital circuits from normal services provides design
economy, since hardening-type construction methods can be focused on only
those specialized areas. Judicious use of steel conduits, pipes, and ducts associated with life-safety systems provides increased likelihood of survivability for
these systems, and a greater likelihood that emergency systems will remain operational post-event to assist rescuers in the evacuation of the building, and/or
provide continuity of critical business services.
Although each design situation is unique, it is possible to provide some general guidance regarding blast-resistant design of equipment and systems. This
section will discuss choosing appropriate design goals, specific points of concern for the different loads that can be imposed, and some guidelines regarding
select types of equipment and systems. However, good engineering practice and
judgment are the best approach for these situations, given the amount of uncertainty present in this field.
334
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
13.3.1 Level of Protection
A critical step in the design of building systems for blast load is to determine
the goal of hardening. In general, there are two levels of performance normally
considered in system blast-resistant design:
1. Life Safety: Also called position retention. This design level is intended to
keep the equipment in place, with no large pieces or components coming
loose and potentially injuring personnel. The equipment or system is not
expected to remain operational after the event.
2. Continued Operation: As the name implies, the equipment is expected to
remain operational after the event. This might entail resetting the equipment, but no major repairs should be required to maintain basic operational
capabilities.
Continued operation of equipment subjected to blast or shock loads is often
difficult to predict. Seemingly minor failures can lead to a shutdown and loss
of critical systems. In general, only inherently rugged equipment can be considered for continued operation if directly exposed to blast pressures. Less rugged
equipment may be designed to remain operational for reduced pressures due to
protection by shelters or interior location. If the continued operation of a system
is required, it is essential that all aspects of the loading condition be considered,
and testing of the equipment is likely to be necessary.
The life-safety level of design is easier to achieve, as no evaluation of the
working parts of the equipment is required. It is only necessary to analyze the
equipment structure and anchorage. Note, however, that the analysis includes all
external components, including the sheet metal shell of the equipment, if any, to
ensure that they remain attached to the equipment. Large internal components
that have the ability to break out of the shell must also be considered. Another
response quantity that must be considered for suspended systems is deflections.
A system may be capable of resisting the pressure or acceleration load on its
own, but large deflections may cause an impact with other equipment and lead
to an anchorage or equipment failure.
13.3.2 Blast Pressures
The best way to protect equipment and systems from the effects of blast is to
locate them in areas protected from direct air blast. If the building envelope is
properly designed to resist blast effects, location of equipment in mechanical
rooms without openings to the exterior of the building will preclude pressure
loading from external events. However, equipment that is located inside protective structures will often be subjected to blast pressure that leaks in through openings in the walls and/or roof. Likewise, properly securing hardened mechanical
rooms, and thus limiting access to authorized personnel, will significantly reduce
the likelihood of internal blast loads on the equipment. However, concentrating
DESIGN CONSIDERATIONS
335
mechanical equipment in a single location will render the overall system more
vulnerable to unauthorized access if proper precautions are not maintained.
Removal of equipment from vulnerable areas is not always possible, and is
likely not feasible for all distributed systems. In these situations, it is necessary to
require design to higher than desired levels of pressure loading. This may entail
providing barrier walls, if not complete shelters, or testing equipment to the anticipated design load. Particularly critical distributed systems may be enclosed in
hardened runways, or redundant systems located elsewhere in the building may
be provided. Obviously, these additional precautions can be extremely expensive, and care should be taken early in the design process to eliminate or reduce
to a minimum the number and types of systems that require special design.
13.3.3 Shock Induced by the Structure
The impact of blast-generated air and ground shock on the structure will produce
motions that are potentially damaging to the building systems. The resulting
structural accelerations can be estimated using the methods outlined later in
this chapter. After the structural motions are calculated, the resistance of the
equipment to shock loads can be checked. Typical values of equipment shock
resistance are given in Table 13.1. One major caveat applies to the use of these
values. In many cases they are based on testing that is decades old. In the interim
many changes have occurred in the design of some equipment. The tabulated
values may no longer be applicable, and caution should be used in applying
them to modern equipment. Significant testing is often required for equipment
used in the military and the nuclear industry, and manufacturers may be able
to provide capacities for specific pieces of equipment. The nuclear industry
Seismic Qualification Utility Group (SQUG) has a large database of equipment
response during earthquakes, but these data are not generally available to
the public.
A common practice is to specify design loads for the equipment in terms of an
equivalent earthquake load. This is particularly true where it is desirable that the
actual load source remain confidential. It is perfectly acceptable to give only an
equivalent “g” acceleration value. The acceleration level to be specified can be
determined using the methods detailed in Chapter 7 or a more detailed calculation that considers the stiffness and mass distribution in the structure. However,
there are several caveats that must be observed.
First, unless a dynamic analysis of the building is performed and the floor
accelerations specified as time histories, the frequency content of the input motion to the equipment is lost. Second, it is not correct to specify a design to,
for example, 0.75g using the 2006 International Building Code (IBC). The IBC
method is used to determine earthquake loads for anchorage design and equipment qualification. A base ground acceleration is used that is then modified to
account for location within the building, natural frequency of the equipment,
ability of the equipment to undergo nonlinear deformations without failure, and
the importance of the equipment.
336
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.1
2007)
Selected Equipment/System Fragility Level Guidelines (ASHRAE
Accelerations (g)
Equipment/System Type
No Damage Probable
Minor Damage Probable
Pumps up to 100 hp
4.5
8
Control Panels
3.5
8
Fans up to 100 hp
4
9
Welded steel piping (18 in. dia max, Sch
30 and up)
4
8
Air-handling units
4.5
10
Duct, steel
4.5
9
Motor control centers
2.5
N/A
Uninterruptible power supplies
3
N/A
Cable trays (up to 36 in.)
4.5
8
Chillers
Centrifugal
4
N/A
Absorption
4.5
N/A
Screw
3.5
N/A
Air-cooled
3.0
N/A
3.5
N/A
Boilers
There are several contributing factors as to why it is not good practice to
apply this methodology to in-structure shock loads. First, earthquake loads on
equipment depend not only on the size of the earthquake, but also on the vertical
location within the building. If the floor accelerations have been determined, and
that is what is being specified, it must be made clear that the factor for vertical
location (an increase in load of as much as three times) should be disregarded in
the calculation. It is also not clear if the other factors are applicable. However, the
most critical point is that many parts of the country are exempt from providing
seismic restraint to equipment regardless of the specified acceleration level. For
these reasons it is best to provide an acceleration level, either ground or floor
level; require use of the building code load equation as appropriate; and explicitly
state that no exemptions are permitted.
Shock Isolation: Isolation of equipment to reduce the effects of shock loading has been used successfully on many projects. Isolation systems work on
the principle that if the loading and isolation system’s natural frequencies vary
greatly, the isolated equipment will undergo motions much less severe than the
base structure. The amount the motion is reduced depends upon the ratio of the
forcing frequency to the natural frequency of the isolation system (frequency
ratio), as illustrated in Figure 13.1. Thus, to achieve a high level of reduction,
the isolation system should have a much lower natural frequency than the input
DESIGN CONSIDERATIONS
337
Figure 13.1 Transmissibility of Motion for Isolated Systems with 5% Damping
motion. However, this leads to a soft isolation system (low stiffness) and potentially high static deformations and problems with coupling between displacement
and rotation deformation modes. A balanced system using a properly selected
isolator, additional weight (inertia mass), and possibly supplemental damping
can help to alleviate the problems associated with soft isolation systems.
There are several types of isolators that are commonly used, including
steel springs, wire coils, air-springs, elastomeric mounts, and damping devices
(Figure 13.2). Most of the isolators can be used in either direct-mounted configurations, where the inertia mass and equipment are placed on top of the isolators,
or in a pendulum configuration where the load is hung from the isolators. In
either case, the systems are designed to carry the gravity load and accommodate
the deformations expected during a shock load. The equipment thus remains
relatively motionless while the supporting floor undergoes large motions, with
the isolation system deforming while maintaining vertical load capacity. Shock
isolation systems are available from a number of manufacturers who provide
design services along with the isolation systems. The project engineer must
determine the level of isolation required and the input motions, while the
manufacturer will choose the correct isolator to achieve the desired results.
13.3.4 Equipment/System Anchorage
The anchorage of equipment to the structure must be checked for all loading
cases and desired levels of performance. The procedure is relatively simple: Analytically apply the load to a mathematical model of the equipment, calculate
338
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
CMM Plate
Inertia Base
Kinetics
PSS Mount
(a)
(b)
Figure 13.2 Examples of Spring Isolators for Shock Isolation: (a) Spring isolator in
a pendulum configuration (Kinetics Noise Control, Inc.); (b) Wire rope isolators (VMC
Group, Inc.); (c) Air-spring isolator (Kinetics Noise Control, Inc.); (d) Elastomeric isolators (Kinetics Noise Control, Inc.)
the forces at the support location, and design the anchorage to accommodate the
maximum forces. The equipment is usually assumed to be rigid for the purpose
of calculating anchorage forces, since the dynamic effects are not considered.
If desired, a more detailed model that considers the mass and stiffness of the
equipment may be used. The load is applied at the center of the exposed surface for pressure loads, and at the center of mass for accelerations. It may also
be necessary to account for nonsymmetric loading and/or mass by varying the
load application point. In addition, the load should be considered to act in all
DESIGN CONSIDERATIONS
(c)
(d)
Figure 13.2 Continued...
339
340
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
possible directions. Note that this is not simply all directions, as in seismic design, since the explosive location may be restricted and therefore limit the load
direction. All seismic design guidelines should be observed regarding limitations
on the type of anchors. For example, power-actuated fasteners should not be used
unless specifically qualified, and undercut anchors are required for equipment
over 10 hp that is not vibration-isolated. In addition, the structural component to
which the equipment is mounted must also be capable of carrying the anchorage
force. Details on the anchorage calculations are provided in the design examples.
13.3.5 Placement of Critical Systems Equipment and Control Stations
Unsecured areas such as the lobby, loading dock, mailroom, garage, retail, and
other public access areas must be separated from the secured areas of the building, particularly areas of a building housing switchgear, computer server data
centers, centralized command centers, and machinery plant control and monitoring stations. Ideally, unsecured areas are placed exterior to the main building or
along the edges of the building. Occupied areas or areas providing emergency
or critical functions should not be placed immediately adjacent to the lobby, but
should be separated by a substantive buffer area such as a storage area or corridor.
From an operational security standpoint, it is important to restrict and control access to air-intake louvers, mechanical and electrical rooms, telecommunications
spaces, and rooftops by means of such measures as visitor screening, limited elevator stops, closed-circuit television (CCTV), detection, and card access-control
systems.
13.3.6 Staffing and Building Operations
How building staffing (i.e., managers, maintenance crew, boiler-men, security
personnel, attendants, etc.) is determined can make a substantial difference in
minimizing loss of life and minimizing injury, as well as improving the ability of a facility to sustain continuity of critical services in the event of a blast.
Determining staffing is integral to the design-development process of building
systems. For instance, automation and systems monitoring rely on certain assumptions regarding staffing. Increased use of centralized equipment/systems
health-determining methods, such as computerized condition-based monitoring
(vibration, temperature, differential pressures, etc.) significantly reduces the need
for periodic roving manned inspections.
An excellent model for effective staffing can be taken from the military. U.S.
warships are designed to absorb a certain number of strikes to minimize loss of
life, ensure continued reliable maneuvering and control (survival of steering gear,
propulsion systems, etc.), and ensure reliable operation of both defensive and
offensive systems. Such vessels rely on an architecture of self-sufficiency using
highly trained crews who are assigned damage control functions in addition to
their normal duties (e.g., cook and fireman). As with warships in battle, buildings
DESIGN CONSIDERATIONS
341
in the face of a civilian mass casualty incident need be self-sufficient, since it is
unlikely that first responders will be able to respond promptly, if at all.
Many elements of the military model can be easily adopted by the civilian
community, and much can be gained by investing in certain added training and
emergency duty assignments for critical personnel. Although using better-trained
personnel may increase operational costs, strategic optimization of variable resources will offset cost increases. For instance, since such personnel are better
trained in building systems, they could also perform certain maintenance and
repair functions routinely outsourced to specialists—serving not only to save
money but likely to reduce downtime of building systems.
The availability and familiarity of selected record drawings to such designated
personnel (i.e., electrical one-line, list of feeders and mains, equipment location
drawings, etc.) and the periodic drilling of personnel promote familiarity of systems while also identifying weaknesses in capability.
13.3.7 Construction of Hardened Spaces
Spaces dedicated to housing critical equipment should be constructed to withstand the effects of blast and fire in unsecured adjacent spaces. Historically, the
preferred material for explosion-mitigating construction has been cast-in-place
reinforced concrete. Reinforced concrete has a number of attributes that make it
the practical construction material of choice. The use of the new tough polymer
spray-on coatings on exposed concrete surfaces, as commonly used to protect the
R
R
beds of commercial pickup trucks (e.g., Line-X , Rhino .) checks the release
of airborne debris in the event of blast. Outfit and furnishings located in secure spaces should be selected to minimize fire load, while electrical equipment
should be mounted on resilient foundations to minimize the effects of shock.
Depending upon the function and equipment located in the secure space, consideration should be given to automated conflagration control systems, such as CO2
deluge systems, the use of fire-retardant cable and cable systems, and the use
of intumescent coatings. Positive pressure systems and/or gas-tight construction
should be considered in critical spaces to control air flow and thwart and control
undesired migration of contaminants.
13.3.8 HVAC and Plumbing Systems
Equipment that is deemed essential should be placed in secure locations. That
is obviously not possible for all system components, as they must be distributed
throughout the building. Because of that, design of HVAC/plumbing systems
usually focuses on position retention (adequate anchorage). Systems that service
areas that must remain functional after an event—for example a safe haven or
control center—should be served by redundant systems or have the equipment
and delivery systems placed in secure areas and hardened runs. The high cost
associated with these countermeasures necessitates a thorough evaluation of the
expected threat and the consequences of system failure.
342
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Air-intake locations should be located as high up in the building as practical,
to limit access and reduce the blast pressure level. Generally, air intake should
be located in secure areas. Systems that serve public access areas such as mailreceiving rooms, loading docks, lobbies, freight elevators/lobbies should be isolated and provided with dedicated air-handling systems capable of 100% exhaust
mode. Air-intake locations and fan rooms should be coordinated with the security
surveillance and alarm system so that surveillance is provided and unauthorized
access can be responded to by the security force.
Building HVAC systems are typically controlled by a building automation
system, which allows for quick response to shut down or selectively isolate
air-conditioning systems. These systems should be coordinated with the smokecontrol and fire-alarm systems to allow for automated, proper response to fires.
Access to building mechanical areas (e.g., mechanical rooms, roofs, elevator equipment access) should be restricted and monitored. These areas provide
access to equipment and systems (e.g., HVAC, elevator, building exhaust, and
communication and control) that could be used or manipulated to assist in smoke
evacuation after a blast or during a chemical, biological, or radiological (CBR)
attack. Additional protection may be provided by including these areas in those
monitored by electronic security and by eliminating elevator stops at the levels
that house this equipment. For rooftop mechanical equipment, ways of restricting (or at least monitoring) access to the roof that do not violate fire codes should
be considered.
Particular care should be taken to ensure that HVAC equipment and attached
ducts/pipes can accommodate any expected deformations. This generally means
using flexible connectors between duct/pipe and equipment, an example of which
is shown in Figure 13.3. In addition, distributed systems often cross areas with
Figure 13.3 Example of Flexible Connectors Near Equipment Attachment Points
(Courtesy of Kinetics Noise Control, Inc.)
DESIGN CONSIDERATIONS
343
large differential motions, including stories with different floor motions and
building joints. Attachments at these locations should also provide adequate deformation capacity to prevent damage to the duct or piping. Possible configurations and options for allowing displacements while retaining functionality at a
pipe penetration through a floor are shown in Figure 13.4.
Locations where building systems intersect with the building envelope
can require additional considerations. Air intakes can be fit with blast valves
(Figure 13.5) that allow for air flow under normal conditions but close when
excess pressure is detected. These systems can be either passive or active.
Penetration of piping or conduit through the wall can allow blast pressures
(a)
(b)
(c)
Figure 13.4 Examples of Pipe Floor Penetration Supports That Allow Differential
Movement: (a) Pipe riser configurations; (b) Riser guide; (c) Riser isolator (Courtesy
of Kinetics Noise Control, Inc.)
344
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
(a)
(b)
(c)
Figure 13.5 Conceptual Operation of a Blast Valve: (a) Normal flow; (b) Excessive
pressure from outside; (c) Excessive pressure from inside
to enter the building and/or damage the mechanical system. Various available
schemes, including sleeved openings, special accumulators (Figure 13.6) that
allow for pressure buildup and gradual release, and flexible connections can
protect both the piping and the building interior.
Fire protection systems may be considered a subset of piping systems and
designed using the same general principles. However, while the failure of water piping systems is unlikely to jeopardize the building occupants, leaking or
inoperability of fire protection piping (and pipes containing hazardous material)
is potentially life threatening. For this reason it is recommended that additional
caution be exercised when designing these systems.
13.3.9 Electrical Systems
Electric power supply, command, control, and communications systems are
essential to all building systems and a proper threat-assessment matrix evaluation is central to the reliability and survivability of such services. A proper
threat-assessment evaluation consists of identifying critical system components,
dependencies, vulnerabilities, and redundancies, and mapping such items in a
DESIGN CONSIDERATIONS
345
Figure 13.6 Conceptual Design of Accumulator (After U.S. Department of the Army,
1986)
documentable “matrix.” Deliverables resulting from such an effort frequently
result in a Failure Mode and Effects Analysis (FMEA).
For electric systems, certain methods or design techniques are frequently used
in guarding against unconventional threats, including:
r Vertical and horizontal zonal definition of equipment and feeds to provide
redundant paths and feeds
r Self-healing power supply and distribution network designs comprising
multiple, segregated power feeds to minimize human-element intervention
r Command, control, and communication (C3 ) services combining both distributed and centralized control techniques to further minimize the need of
human resources
r Tactical location of critical equipment and personnel
r Co-location of critical equipment in hardened/controlled-access spaces
r Application and integration of automation systems to monitor and orchestrate power, control, and communication systems healing actions
r Appropriate personnel matrix and personnel training to increase the likelihood of survivability/mission continuity, while reducing variable resource
costs
Utility power transformer(s) and switchgear should be located interior to the
building if possible, and in secure spaces. For larger buildings, multiple transformers, separated from each other, enhance reliability should a blast damage
one transformer or its associated equipment. In circumstances where a redundant utility transformer arrangement is not provided, consideration should be
given to locally stocking a replacement transformer and associated equipment
when replacement lead times may be unacceptably long.
346
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
13.3.10 Lighting Systems
Lighting system design for blast protection is similar to electrical system design. The same principles of separation, zoning, emergency power supply, and
so forth, all apply to lighting. One recent development that should be considered
in lighting design is the use of light-emitting diode (LED) technology. More
frequently recognized in the electronic equipment arena, where LEDs are used
for equipment pilot and status lights, LED technology has now evolved to offer
very practical and cost-effective design alternatives in applications where incandescent and fluorescent lamps presently dominate. With respect to susceptibility
to blast damage, LEDs stand alone as a clear leader. In addition, LED lighting reduces the environmental impact of the lighting system and offers reduced
maintenance and increased life.
LED lighting methods do not depend on moving parts or parts that are motionsensitive, as do incandescent and fluorescent lamp technologies (filaments, standoffs, glass, etc); that is, LEDs indeed live up to their designation as “solid-state,”
and their elemental constituents, such as device leads, are solidly embedded into
inert and tough structural polymers. Their substantively smaller sizes and physical profiles minimize concerns related to weight and moment and susceptibility
to blast damage, while also eliminating the need for support devices/equipment
(e.g., ballasts) and the weight/room required for these other support devices.
13.3.11 Other Systems/Considerations
Emergency Power System An emergency generator should be available to provide an alternate source of power if utility power becomes unavailable to support critical life-safety systems such as alarm systems, egress lighting fixtures,
exit signs, emergency communications systems, smoke-control equipment, and
fire pumps. In cases where continuity of critical business missions is necessary, the emergency generator system can also support these missions. Practically, diesel-fueled emergency generator systems are preferred over other conventional fuels (i.e., gasoline, propane, etc.) because of fuel availability, lower
flash point of diesel fuel, and greater power supply capacity from a relatively
small footprint. Economically, maintenance costs are substantially lower for a
diesel-driven emergency generator system when contrasted to other forms of
emergency power, including fuel cells and battery-based systems.
Emergency generators typically require large louvers to allow for ventilation
and cooling of the generator. Care should be taken to locate the generator so that
these louvers are not vulnerable to attack or are fit with blast valves. A remote
radiator system could be used to reduce the louver size and thus reduce air-intake
requirements.
Ideally, emergency power-distribution feeders are contained in hardened enclosures, or encased in concrete, and configured in redundant routing paths to
minimize vulnerability and thus enhance reliability. Emergency distribution panels and automatic transfer switches should be located in hardened spaces isolated
DESIGN CONSIDERATIONS
347
from the normal power system. Emergency lighting fixtures and exit signs along
the egress path could be provided with integral battery packs to provide closetransition services (uninterrupted silhouette illumination) between the time the
utility system fails and the emergency generator system assumes loading.
Emergency Generator System and Fuel Storage A nonexplosive fuel source,
such as diesel fuel, should be considered for emergency generators and dieseldriven fire pumps. The preferred philosophy is to locate all support equipment,
including the diesel fuel day tank, in the same space as the emergency generator,
as a means of minimizing system dependencies. Consideration should be given
to locating the emergency generator day tank in a fire-hardened space adjacent to
the emergency generator space. Day tanks should include sufficient fuel to power
the generator for a minimum of eight hours.
All control equipment, including starting provisions (i.e., air start, battery
start, hydraulic start, etc.), emergency generator switchgear, cooling water tank,
local controls, and so forth, should co-locate with the emergency generator.
Diesel storage tanks that augment the day tanks should be located away from the
building in secure, fire-rated, hardened structures. Fuel piping within the building
should be located in hardened enclosures, and redundant piping systems could
be provided to enhance the reliability of the fuel distribution system.
Fire Control Center A centralized Fire Control Center (FCC) should be
provided to monitor alarms and life-safety components, operate smoke control
systems and fire doors, communicate with occupants, and control the firefighting/evacuation process. Consider providing redundant Fire Control Centers
remotely located from each other, to allow system operation and control from
alternate locations. The FCC should be located near the point of firefighter
access to the building. If the control center is adjacent to the lobby, separate
it from the lobby using a corridor or other substantive buffer area. Provide
hardened construction for the Fire Control Center. Further, the FCC can be
co-located with other critical automated building service functions, such as
with the machinery plant control and monitoring systems. The proliferation of
computer-based monitoring and control systems permits much flexibility in this
regard, particularly for collecting multiple, monitored systems into a single space
through building intranet cabling. If configured, a simple laptop computer can
be used to access and control monitored systems from many possible locations.
Emergency Elevator System Elevators are not used as a means of egress from
a building in the event of a life-safety emergency event, as conventional elevators
are not suitably protected from the penetration of smoke into the elevator shaft.
An unwitting passenger could be endangered if an elevator door opened onto
a smoke-filled lobby. Firefighters may elect to manually use an elevator for
firefighting or rescue operation. A dedicated elevator, within its own hardened,
smoke-proof enclosure, could enhance the firefighting and rescue operation after
a blast/fire event. The dedicated elevator should be supplied from the emergency
348
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
generator, fed by conduit/wire that is protected in hardened enclosures. This
shaft/lobby assembly should be sealed and positively pressurized to prevent the
penetration of smoke into the protected area.
Smoke and Fire Detection and Alarm System A combination of earlywarning smoke detectors, sprinkler-flow switches, manual pull stations, and audible and visual alarms provides quick response and notification of an event.
The activation of any device will automatically start the sequence of operation of smoke control, egress, and communication systems to allow occupants to quickly go to a safe area. System designs should include redundancy
such as looped infrastructure wiring and distributed intelligence such that the
severing of the loop will not disable the system. Install a network of firealarm systems consisting of distributed intelligent fire-alarm panels, such that
each panel can function independently, process alarms and initiate sequences
within its respective zone, while at the same time communicating with other
panels.
Smoke-Control Systems Appropriate smoke-control systems maintain smokefree paths of egress for building occupants through a series of fans, ductwork,
and fire-smoke dampers. Stair pressurization systems maintain a clear path of
egress for occupants to reach safe areas or to evacuate the building. Smokecontrol fans should be located higher in a building, rather than at lower floors,
to limit exposure/access to external vents. Vestibules at stairways with separate
pressurization provide an additional layer of smoke control.
General Announcing Communication Systems A general announcing system
facilitates the orderly control of occupants and evacuation of the danger area
or the entire building. The system is typically zoned by floor, by stairwell, and
by elevator bank for selective communication to building occupants. These systems may be integrated with other building announcing and communication systems. Emergency communication can be enhanced by providing extra emergency
phones separate from the telephone system, connected directly to a constantly supervised central station, and an in-building repeater system for police, fire, and
EMS (emergency medical services) radios.
13.4 LOADING CALCULATION
Nonstructural components can be subjected to several different load types due to
blast. This section details the procedures used to calculate the blast loads, including direct loading, leakage through openings, and propagation through ducts. In
addition, calculation of the building motions for design of systems is addressed,
as is the load due to accidental blastlike loads such as short circuits in electrical
equipment.
LOADING CALCULATION
349
13.4.1 Blast Pressure
Building systems can be subjected to blast pressure from three sources: direct
external blasts, direct internal blasts, and indirect pressures. Equipment located
external to the building shell will be exposed to direct blast pressures. These
pressures are characterized by a short rise time, high peak pressure, and relatively
fast drop of pressure. In contrast, pressures resulting from internal blasts are
confined by the walls and consequently have longer duration. The calculation of
blast pressures due to both external and internal blasts is covered elsewhere in
this handbook (see Chapter 7).
Leakage through Openings Equipment housed in shelters or in rooms with a
direct connection to the building exterior will experience blast pressures from
leakage through openings. These loads may be applied directly to the equipment with connections to the outside, such as air handlers with external intake/exhaust, or indirectly as the pressure fills the room. The calculation of
interior pressure loads due to leakage is complex and usually involves many
assumptions.
One procedure, presented below, incorporates the prominent assumption
that the openings are too small to allow a shock front to develop, and that the
pressure inside the room increases uniformly. The method calculates the average
pressure in the room. The pressure near the opening will be somewhat larger
than the average, but the approximation of uniform pressure within the space is
adequate for design.
The interior pressure loading history is calculated in a step-by-step manner.
The procedure is as follows:
1. Calculate the external pressure-time history.
2. Divide the duration (to ) of the external load into equal intervals of length
δt equal to approximately to /10 or to /20.
3. For each time interval, calculate the change in internal pressure over the
time step, using
δ Pi = C L Ao Vo δt
where CL is the leakage pressure coefficient (Figure 13.7) (Departments of
the Army, the Navy and the Air Force 1990), taken as negative when P is
negative; Ao is the area of the opening; and Vo is the room volume. The
internal pressure at the end of the time interval is then calculated from the
pressure at the beginning of the step as
Pi = P − δ Pi
4. Repeat until the internal pressure drops to the initial ambient pressure.
350
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Figure 13.7 Leakage Pressure Coefficient (After Departments of the Army, the Navy
and the Air Force, 1990)
Example: Calculate the internal pressure-time history for an equipment penthouse with a small opening in one wall.
Given:
Ao = 4 ft2 (2 × 2 opening)
Vo = 1000 ft3 (10 × 10 × 10 penthouse)
Assume peak external pressure at time 0 equal to 3 psi with to = 50 ms.
Solution:
Use δt = to /10 = 50/10 = 5 ms
Perform calculations in a tabular format. Results are shown in Figure 13.8.
LOADING CALCULATION
351
(a)
(b)
Figure 13.8 External and Internal Pressure-Time Histories for Leakage Example: (a)
External pressure-time history; (b) Internal pressure-time history
Propagation of Pressure through Ducts Blast pressure will propagate through
ducts into the interior of a building. The pressure entering the duct, and the
change in pressure at different branching configurations, can be calculated using Figure 13.9. These figures are valid for peak overpressures less than 50
psi, which is appropriate for most civilian structures. In addition, 90◦ bends
in the duct reduce the pressure by approximately 6% each. Thus, after each
352
T (ms)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
...
185
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
P (psi)
Pi (psi)
P-Pi (psi)
CL
δPi
3.00
2.70
2.40
2.10
1.80
1.50
1.20
0.90
0.60
0.30
0
0
0
0
0
0
0.17
0.31
0.43
0.53
0.60
0.65
0.68
0.69
0.69
0.67
0.63
0.59
0.56
0.53
3.00
2.53
2.09
1.67
1.27
0.90
0.55
0.22
−0.09
−0.39
−0.67
−0.63
−0.59
−0.56
−0.53
8.5
7.17
5.91
4.73
3.61
2.55
1.56
0.62
−0.26
−1.10
−1.89
−1.78
1.68
−1.58
−1.49
0.17
0.14
0.12
0.09
0.07
0.05
0.03
0.01
−0.005
−0.02
−0.03
−0.04
−0.03
−0.03
−0.03
0
∼0
—
—
—
bend, the peak pressure, PT , can be calculated from the pressure before the
bend, Po , as
PT = 0.94Po
There is also pressure loss along the length of smooth ducts. However, the
calculation is not suggested for the typical ranges and charge weights assumed
in this chapter, because it is only slightly conservative to assume no loss along
straight portions of ducts. This assumption will be overly conservative for either
long duct runs (over 50 feet) or short ranges. Further details can be found in TM
5-855-1 (U.S. Department of the Army 1986) if more detailed calculations are
desired.
13.4.2 In-Structure Shock
In-structure shock refers to the structural motions that result from a blast. Although equipment located within the structure may not be subjected to direct or
indirect blast pressures, it will experience the resulting structural accelerations.
As a result, building systems not exposed to pressure loads must still be designed
to resist the in-structure shock.
Structural motions due to blast arise from three sources: air-induced ground
shock, direct-induced ground shock, and air shock. Air-induced ground shock
is the result of the pressure wave acting on the ground surface and producing
deflections and hence motion on the ground. Direct-induced ground motions are
due to direct transfer of energy to the ground, typically due to buried or surface detonations. Air shock occurs when the blast pressure impacts the building,
LOADING CALCULATION
(a)
(b)
(c)
(d)
(e)
(f)
353
Figure 13.9 Transmitted Overpressures in Ducts (After U.S. Department of the Army,
1986)
imparting structural motions. The effect of all three sources must be considered
when determining the design motions for building systems.
Air-Induced Ground Shock The peak vertical and horizontal accelerations due
to air-induced ground shock are calculated as
AVai = 100Pso (ρC pg)
AHai = A V tan sin−1 C p (12000U )
where AVai and AHai are the air-induced vertical and horizontal peak ground accelerations (g), Pso (psi) and U (ft/ms) are the peak positive incident pressure and
354
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.2 Typical Values of ρ and Cp (After Departments of the Army, the Navy
and the Air Force, 1990)
Soil Type
ρ (lb-sec2 /in4 )
Soil Type
Cp (in/sec)
Loose, dry sand
1.42E-04
Loose and dry soils
7200–39,600
Loose, saturated sand
1.79E-04
Clay and wet soils
30,000–75,600
Dense, dry sand
1.65E-04
Coarse and compact soils
36,000–102,000
Dense, saturated sand
2.02E-04
Sandstone and cemented soils
36,000–168,000
Dry clay
1.12E-04
Shale and marl
72,000–210,000
Saturated clay
1.65E-04
Limestone-chalk
84,000–252,000
Dry, sandy silt
1.57E-04
Metamorphic rocks
120,00–252,000
Saturated, sandy silt
1.95E-04
Volcanic rocks
120,000–270,000
Basalt
2.56E-04
Sound plutonic rocks
156,000–300,000
Granite
2.47E-04
Jointed granite
9600–180,000
Weathered rock
24,000–120,000
Limestone
2.25E-04
Sandstone
2.10E-04
Shale
2.17E-04
Concrete
2.25E-04
shock front velocity, ρ (lb-sec2 /in4 ) is the soil mass density, Cp (in/sec) is the soil
seismic wave velocity, and g is the acceleration due to gravity (32.2 ft/sec2 ). Typical values of ρ and Cp are given in Table 13.2. If Cp /12000U is greater than 1,
take AHai = AVai .
Direct-Induced Ground Shock Direct-induced ground motions depend upon
the charge weight, distance from the structure, and the soil conditions. The peak
vertical ground acceleration from direct-induced ground shock can be calculated
from
A V di = 10,000 W 1/3 Z G2
where W is the TNT equivalent charge weight (lb), and ZG is the scaled ground
distance (ft). The peak horizontal acceleration depends upon the soil conditions
and is
AHdi =
0.5A V
1.0A V
dry soil
wet soil or rock
The arrival time of the direct-induced ground shock at the structure can be
calculated from
t AG = 12,000RG /C P
where RG is the ground distance from the explosion (ft).
LOADING CALCULATION
355
Air Shock Air-shock motions can be calculated in two ways. A simplified
method is presented in this chapter that assumes a rigid structure and is based
on sliding of the foundation. For this method, only the horizontal motions are
considered, as rocking and vertical motions due to air shock are assumed to
be negligible. Optionally, a more sophisticated analysis can be performed that
accounts for the time history of the load and the flexibility of the structural elements. Although this type of analysis is beyond the scope of this handbook, some
general guidelines are presented. For all but the smallest blast loads, nonlinear
behavior of the structural elements, at least locally near the blast location, is expected and must be included in the analysis model. The load should be based on
the design charge weight and standoff and the time history of pressure applied
to the exposed surfaces of the structure. Care must be taken to ensure that the
stiffness, strength, and mass distribution are modeled as accurately as possible.
The general level of detail used in a dynamic, nonlinear analysis for earthquake
loads is appropriate. One exception to this is that the out-of-plane behavior of
walls is critical in determining the structure response to blast loading, and the
simple linear models used for this behavior mode in earthquake analyses are not
adequate. (See Chapters 6, 7, and 9 for additional information.)
The simplified method starts by assuming that the structure has a single degree
of freedom and that the only resistance to motion is the foundation stiffness. The
resulting equation is
F (t) − FR (t) = M AHas (t)
where F is the applied (blast) load, FR is the resistance of the structure foundation, M is the structure mass, and AHas is the resulting horizontal acceleration.
For a shallow footing, the resistance function can be assumed to depend only
upon friction between the foundation and soil, and is calculated as
FR = µN
where µ is the coefficient of friction (Table 13.3), and N is the normal force
acting on the footings. N is a combination of the structure weight and the vertical
load applied to the roof through the blast pressure. Thus N, and subsequently FR ,
varies with time.
Ideally, the equation of motion would be solved at each time step, and the
maximum acceleration extracted from the results. However, since the load is
only slightly dependent upon the motion, movement in the direction of the load
reduces the effective pressure, and the resistance is assumed to be independent
of motion, a practical alternative is available. The peak pressure and minimum
value of N are assumed, and the resulting acceleration is used as the peak value.
This results in a peak air-shock horizontal structural acceleration of
AHas =
Pso A S − µWs
M
where AS is the exposed area of the structure, and Ws is the structure weight.
356
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Table 13.3 Coefficient of Friction between Concrete and Soil (After Departments
of the Army, the Navy and the Air Force, 1990)
Soil Type
µ
Clean sound rock
0.70
Clean gravel, gravel-sand mixture, coarse sand
0.55–0.60
Clean fine to medium or silty medium to coarse sand, silty or clayey gravel
0.45–0.55
Clean fine sand, silty or clayey fine to medium sand
0.35–0.45
Fine sandy silt, nonplastic silt
0.30–0.35
Very stiff and hard residual or preconsolidated clay
0.40–0.50
Medium stiff and stiff clay and silty clay
0.30–0.35
Load Combination The air-induced ground shock and air shock may be assumed to arrive at the structure simultaneously, since both are the result of the
blast pressure. The ground-induced shock wave will arrive at a later time, since
the wave speed through soil is lower than through air. Since only the peak motions are considered, whether or not the motions directly combine depends not
only upon the arrival times, but also upon the actual peak motion times induced
within the structure. The actual time histories of the in-structure motions are
extremely difficult to calculate due to the lack of adequate data on the time
history of the input loads for ground shock and the uncertainty in structural
response.
Whether or not the direct-induced ground shock adds to the air-induced
ground shock and air shock depends upon the arrival time, tAG . If tAG is greater
than the combined arrival time and duration of the blast load (ta + t0 ), then the
air-induced motions can be assumed to dissipate before the direct-induced shock
arrives, and the peak accelerations should not be added. However, if tAG is less
than or equal to ta + t0 , accelerations from all three sources should be combined.
The equations for all three sources of shock were derived assuming a rigid
structure. Based on this assumption, the motion at all levels of the structure will
be identical. However, although structures tend to be very stiff vertically, this
is not true for the horizontal motions. The ground motion will induce horizontal deformations within the structure, resulting in differing levels of acceleration throughout the building height. This variation of in-structure motion with
height has been codified for earthquakes, with accelerations increasing through
the height due to the response of the building. This is not necessarily true for
blast loads, as the short duration does not allow for structural response in the
same manner as earthquake loads. For example, an analysis of a typical threestory steel building subjected to a 10g ground shock load of duration 50 ms, and
accounting for nonlinear structural response, results in absolute accelerations of
10g at the base, 2g at the second story, and 1g at the roof. Relative displacements
between floors are less than 1 inch. Of course, this is only an example, and the
response in other buildings may be vastly different, but it does illustrate the effect
LOADING CALCULATION
357
of the short load duration on building response and the difference with response
to earthquake loads. Thus, great care must be taken when considering the effects
of location within the building, and it is conservative to assume that the ground
accelerations are present at all levels.
The simplified method for calculating air-shock motions does not allow for
differentiation based on location within the structure. Thus, the vertical accelerations in the structure, assuming simultaneous application of the direct- and
air-induced ground shock, are calculated as
A V = AVai + AVdi
and the horizontal ground accelerations are
A H = AHai + AHdi + AHas
Example
Given: Loose, dry sand
ρ = 1.42E − 04
C p = 7200 − 39,600
µ = 0.45
U = 1.4
1000 − lb TNT surface burst @ 100
W = 1000 lb
Pso = 5.5 psi
Z G = 10
Building: 100 wide × 40 tall, assuming typical density gives
Ws = 3,000,000 lb
M = 7764 lb − s2 /in
Solution:
= 5.14g
AVai = 100Pso (ρC pg) = 100(5.5)/[(.000142)(23,400)(32.2)]
AHai = A V tan
sin−1 C p (12000U
=
5.14g
)
AVdi = 10,000 W (1/3) Z 2 = 10,000 (1000)(1/3) (10)2 =10g
G
AHdi = 0.5A V di = 5g
(5.5)(1200)(480) − (0.45)(3,000,000)
Pso A S − µWs
=
= 0.61g
AHas =
M
7764
The total in-structure motions are
A V = AVai + AVdi = 5.14 + 10 = 15.14g
A H = AHai + AHdi + AHas = 5.14 + 5 + 0.61 = 10.75g
358
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Note that these accelerations are much higher than typically seen in seismic events. For earthquakes, the largest accelerations considered in the lower
48 states for nonstructural component design in the 2006 International Building
Code (International Code Council 2006), not including the effect of component
flexibility, are approximately 5g horizontally and 0.6g vertically.
Anchorage Force Calculation The anchorage forces are calculated by applying
the shock load at the equipment center of mass. The applied force is equal to
the shock acceleration multiplied by the equipment mass. The force should be
applied in all directions, unless the possible blast locations limit the shock direction. Usual practice is to also consider eccentricity of the center of mass, as well
as the actual attachment point locations.
Seismic design principles require that the applied load be further modified
based on the flexibility of the equipment and the amount of nonlinear behavior
that can be tolerated before damage. The flexibility component is intended to
magnify the forces if the equipment will move significantly relative to the floor
(increasing the effective force). For two pieces of equipment of identical mass,
the more flexible equipment will undergo larger deformations and experience
larger accelerations at its center of mass. Another factor that can modify calculated forces recognizes that most equipment can undergo some nonlinear behavior without sustaining significant damage. Since the analysis used to evaluate the
equipment and anchorage assumes linear behavior, a lower effective force will
be transmitted during the actual event. The amount of the reduction that is used
depends on the ability of the equipment to accept nonlinear deformations without damage. However, no formal guidance exists regarding the applicability of
these factors for blast loading, and great care should be taken if they are to be
applied for blast analysis.
Example: Motor Control Center
Equipment data (Figure 13.10):
W
A
B
H
= 2800 lb
= 80 in
= 20 in
= 92 in
Load Data: Use values from previous examples.
A V = 15.14g
A H = 10.75g
Pso = 5.5 psi
Shock calculation:
r Apply the loads AV and AH at the center of mass (assume at center in plan
and midheight). Account for 5% eccentricity in any direction.
(a)
(b)
Figure 13.10 Example Calculation Equipment Layout: (a) Motor control center; (b)
Anchorage layout
360
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
r Calculate the anchorage forces assuming that static loads equal to AV W and
AH W are applied simultaneously. The horizontal load can be applied in any
direction with the maximum anchorage forces governing.
The applied loads are:
Vertical = 15.14g ∗ 2800 lb = 42,392 lb (uplift or compression)
Horizontal = 10.75g ∗ 2800 lb = 30,100 lb
The anchorage forces with the horizontal load applied in the short direction with
zero eccentricity are:
Uplift = (42,392 − 2800)/4 + (30,100)(46 in)/[2(20)] = 44, 513lb
Compression = (42,392 + 2800)/4 + (30,100)(46 in)/[2(20)] = 45, 913lb
Horizontal = 30,100/4 = 7525lb
and the resulting maximum anchorage forces considering eccentricity are:
Vertical = 47,632 lb uplift
Horizontal = 8455 lb
Note that these are considerably higher than 1/4 of the applied loads due to
the location of the load application and the assumed eccentricities. In particular,
the vertical forces are greatly increased due to the overturning moment imposed
by the horizontal force.
These anchorage load values exceed the capacity of 11/4 anchors by approximately 25%. Adequate anchorage would require multiple anchors at each support
location. This compares to one 1/2 anchor required at each support location in
an area of moderate seismic activity (UBC, Zone 3) (International Code Council
1997).
Blast Calculation: Three basic approaches can be used to apply blast loads to
equipment. First, the peak (or average) blast pressure can be applied as a static
force. This is an extremely conservative approach for most equipment, as it ignores the short duration of the applied load. The second approach is to consider
the blast load to be impulsive in nature and to use a simplified method to determine the maximum response. A drawback to this approach is that the dynamic
characteristics, mass and stiffness, of the equipment must be known, which is
usually not the case. A third alternative is to explicitly calculate the time history
response of the equipment to the blast load, but this approach is rarely justified
due to the—at best—approximate nature of the input parameters and the requirement to know the equipment mass and stiffness distribution.
The first two approaches will be considered in this example.
Method 1 – Static Load Application
LOADING CALCULATION
361
r Apply the pressure as a static load distributed evenly across the exposed
face.
r Calculate the anchorage forces for the given loads. The horizontal load can
be applied in any direction, but for symmetric equipment, applying the load
along each principal axis is usually sufficient.
The equivalent applied forces are:
Strong axis = 92 in ∗ 20 in ∗ 5.5 psi = 10,120 lb
Weak axis = 92 in ∗ 80 in ∗ 5.5 psi = 40,480 lb
The resulting maximum anchorage forces are:
Strong axis loading vertical = 5119 lb (uplift)
Strong axis loading horizontal = 2530 lb
Weak axis loading vertical = 185,508 lb (uplift)
Weak axis loading horizontal = 10,120 lb
The anchorage forces generated by these loads are clearly beyond the capacity
of a cost-efficient anchorage system. Several simplifying assumptions were made
in this calculation. In addition to the static application of the load, no load was
applied to the top, side, or rear surfaces. Loads applied on the top and rear face
will reduce the anchorage forces.
The exposed front face of the motor control center must be able to resist the
5.5 psi load. The ability of most sheet metal to sustain such a load is doubtful
at best, with maximum duct pressures approaching 1 psi. Thus, although this
example demonstrates the procedure for calculating the anchorage forces to a
load of 5.5 psi, the equipment would have to be protected through barriers or
relocation to reduce the load to a level that could be resisted.
Method 2 – Impulsive Load
A typical natural frequency for lateral motion of a motor control center might
be 0.5 sec. For this example, that is a lateral stiffness of roughly 1000 lb/in. From
Figure 13.11 the dynamic amplification factor for a ratio of
0.050
t0
=
= 0.1
T
0.50
is D = 0.35. This leads to a maximum total horizontal reaction force of
Htotal =
3542 lbs
0.35(5.5)(92)(20)
=
0.35(5.5)(92)(80)
14,168 lbs
for the strong and weak directions, respectively. This calculation yields horizontal support loads of 885 and 3542 lbs, respectively. Similarly, the vertical forces
362
BLAST-RESISTANT DESIGN OF BUILDING SYSTEMS
Figure 13.11 Dynamic Amplification Factor for Triangular Impulse Loads
due to blast are D times the static values. This results in
V =
2246 lbs
65,382 lbs
The vertical loads (uplift) are still very large, but are likely able to be resisted
with proper anchor design, certainly if direct attachment to steel is possible.
Stiffer equipment (smaller T) will have larger dynamic amplification factors,
while more flexible equipment will see greater reductions in force.
13.5 SUMMARY
Presently, there is little guidance and few standards for blast-resistant design of
building systems for civilian structures. Although there are several efforts currently under way, familiarity and selective adoption of methods used in other
arenas, such as used by the DoD in hardening combatant-type systems, can be
economically and practically exported to the civilian building theater. Likewise,
adoption of new and available technologies, such as LED lighting, can provide
reasonable stopgap measures.
This chapter has attempted to highlight the main areas of concern, provide
basic guidance as to the issues to be addressed, and demonstrate computational
techniques where appropriate. However, given the lack of recognized, standardized design procedure, it is imperative to recognize that, as in all engineering,
there is no substitute for common sense and engineering judgment. The
REFERENCES
363
principles of design remain unchanged, despite the lack of exact knowledge
regarding the loading and capacity for resistance.
REFERENCES
American Society of Heating, Refrigerating and Air-Conditioning Engineers. 2007.
ASHRAE Handbook: Applications. Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers.
Departments of the Army, the Navy and the Air Force. 1990. Structures to Resist the
Effects of Accidental Explosions, rev. 1 (Department of the Army TM 5-1300, Department of the Navy NAVFAC P-397, Department of the Air Force AFM 88-22).
Washington, DC: Departments of the Army, the Navy and the Air Force.
International Code Council. 1997. Uniform Building Code. Washington, DC: International Code Council.
. 2006. International Building Code. Washington, DC: International Code
Council.
National Fire Protection Association. 2005. National Electrical Code Requirements for
One- and Two-Family Dwellings (NFPA 70A). Quincy, MA: National Fire Protection
Association.
U.S. Department of the Army. 1986. Fundamentals of Protective Design for Conventional Weapons (TM 5-855-1). Washington, DC: U.S. Department of the Army.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
IV
Blast-Resistant Detailing
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
14
Blast-Resistant Design
Concepts and Member
Detailing: Concrete
Steven Smith and W. Gene Corley
The next four chapters provide design and detailing guidelines for members
constructed from reinforced concrete, steel, and masonry, as well as members
strengthened using fiber-reinforced polymers (FRP).
The general approach for all of these materials systems is to use two independent approaches to providing the toughness that is critical to blast resistance. The
approaches are informally associated with strength and ductility (which, taken
together, provide toughness). Strength is scaled through the response limits for
each level of protection (LOP). For example, the response limit targets for LOP
IV are elastic response. Thus, members designed to this LOP will be larger and
stronger than those designed to lower LOP (which allow greater displacement).
However, recognizing the universal need for ductility, the detailing recommendations are often constant for all LOP. For the lower LOP, this provides the ductility
necessary to achieve the expected response limits. For higher LOP, this detailing provides a higher margin against the uncertainties inherent to blast-resistant
design.
14.1 GENERAL
14.1.1 Scope
Detailing recommendations for reinforced concrete structures are provided in
this chapter. The nature of high-rise structures in a dense urban environment
introduces significant complexity, including near-field blast loads and intricate
structural members such as transfer girders. For this reason, the methods presented in this section, while generic in nature, are best applied to low-rise structures with moderate standoff distance (say, Z > 3.0).
The seismic provisions in ACI 318-08, Building Code Requirements for Structural Concrete (American Concrete Institute 2008), provide the basis for the detailing, although many references are also made to UFC 3-340-02, Structures
367
368
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
to Resist the Effects of Accidental Explosions (Unified Facilities Criteria 2008).
Chapter 21 of ACI 318-08 contains requirements for design and construction of
structures subjected to earthquake motions. The requirements are scaled to the
severity of earthquake motions and are identified as ordinary, intermediate, and
special moment frames. Special moment frame (SMF) requirements provide the
highest level of performance and are the primary resource for the ACI provisions
referenced in this chapter.
The general intent of seismic detailing is to provide greater ductility and
post-yield strength than conventional detailing. This is consistent with the blastresistant design intent of improving toughness. The Portland Cement Association’s Blast Resistant Design Guide for Reinforced Concrete Structures (Portland
Cement Association 2009) contains additional discussion of structural aspects of
seismic and blast phenomena, as do Hayes (Hayes et al. 2005) and Bilow (Bilow
and Kamara 2007).
14.2 FAILURE MODES
14.2.1 Flexural
Flexure is in many instances intended to be the primary structural mode for members resisting blast loads. This is due to two related concepts: the relatively predictable nature of, and the ductility associated with, flexural response.
Columns subjected directly to lateral air-blast pressures are often the single
most critical design consideration for blast-resistant design. Columns directly
exposed to blast pressures will respond primarily in flexure. In fact, UFC
3-340-02 recommends disregarding the column axial load and analyzing it
solely as a beam. Neglecting the compressive preload is generally a conservative
approach to sizing columns, as long as a separate stability check is performed
that includes the axial load.
Columns may develop a three-hinged mechanism with plastic hinge zones
located at end supports and at an intermediate point along the length of the column. The location of the intermediate plastic hinge can be determined by analysis and depends on the location of the blast source as well as the height and
properties of the column. However, this analysis can be complex and may require assumptions that render results of little use. For most situations, the best
approach is either to conservatively design the column for the potential of a hinge
anywhere along the entire length, or to make general assumptions of intermediate hinge location based on the anticipated blast loading. If the column is in
the far field, the blast pressure will be essentially uniform, and the intermediate
hinge will form near the center of the column (assuming no other structural details that would alter the symmetry of the column response). If the column is in
the near field, and subjected to blast from a package bomb, the blast pressures
will be concentrated at the base of the column, and move the intermediate hinge
downward.
FAILURE MODES
369
Seismic design philosophy as applied in FEMA 350, Recommended Seismic
Design Criteria for New Steel Moment-Frame Buildings (Federal Emergency
Management Agency 2000) employs a target hinge rotation of 0.03 rad (approximately 2◦ ). This hinge rotation can also be associated with reinforced concrete
design (Bilow and Kamara 2007). Potential hinge rotations under blast loading,
however, span a wide range. as described elsewhere in this handbook. The
target seismic deformation magnitude, and overall response intent, would be
best associated LOP II–III. The extreme deformations, structural damage, and
allowance of tension membrane associated with LOP I are beyond the intent of
seismic design.
14.2.2 Diagonal Tension
Diagonal tension shear is a brittle failure mode and must be avoided. Care
must be taken to provide sufficient shear capacity to allow for the full flexural development of the section. The phi-factors provided in ACI 318-08 are
intended to promote preferential failure modes; a greater strength reduction is
applied to shear than flexure. However, the structural analyst has the task of
obtaining the correct member reactions, particularly when conducting singledegree-of-freedom (SDOF) analysis and mapping the results back to the physical
system.
14.2.3 Direct Shear
The primary focus of blast analysis is typically the dynamic flexural response of
members. However, earlier during the blast event, direct shear forces are transmitted to the element supports. The magnitude of the direct shear force may be
significantly greater than the reaction due to dynamic flexural response, though
the resistance to direct shear is also typically greater that diagonal shear resistance.
Direct shear can be an even more brittle mode of response than diagonal tension. Direct shear is the phenomenon of failure at a shear plane, typically at the
face of a support. Direct shear is resisted by shear friction and inclined reinforcement. Specific detailing guidelines for beams and columns are provided later in
this chapter.
14.2.4 Membrane
Slabs and beams that develop extreme deformations from large, directly applied
blast loading may be designed to transition from a flexural response to a tension
membrane response. In addition to careful detailing of the slab or beam, this
mode of response requires careful consideration of the large in-plane forces that
will be transmitted to the surrounding structural elements.
370
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
14.2.5 Stability
Stability is a critical design factor for columns. Designing the column to support
some or even the entire design load after loss of intermediate lateral support is
an indirect means of providing substantial buckling resistance. This equates to
a scenario of the loss of bracing from floor framing, and essentially requires
designing the column for an unbraced length equal to two story heights.
This is not a blast-resistant design method per se, although it clearly relates
to the way the building will resist progressive collapse. However, this is such
an important design consideration that it should be carefully considered, even if
formal progressive collapse design is not utilized. And, this is an essential design
concept whenever a strong-column, weak-beam approach is used.
14.3 DETAILING
14.3.1 General
General procedures that will promote toughness include:
1. Use of concrete with a minimum strength greater than 3,000 psi. This will
help to ensure against brittle shear failure.
2. Concrete with strength greater than 10,000 psi should be used with caution
due to its potentially brittle failure mode in flexure.
3. All concrete should be normal weight. Existing test data and analysis of the
response of reinforced concrete to blast and seismic loading are primarily
based on normal-weight concrete. Thus, for normal-weight concrete, concepts such as strain-rate effects, fragmentation, and pressure-impulse (P-I)
calibration for deformation limits are best understood.
4. The ACI 318-08 design compressive strain in concrete of 0.003 should also
be applied to blast-resistant design, to avoid crushing of concrete and an
associated brittle failure in flexure.
5. Reinforcing bars are recommended to be ASTM A706 (ASTM 2006).
Where ASTM A615 (ASTM 2008) bars are used, limitations as specified
in ACI 318-08, Section 21.2.5, should be applied. The primary concerns
for reinforcing steel are the ratio of yield to ultimate strength and the actual yield versus design yield strength. A large ratio of yield to ultimate
strength is necessary to provide sufficient ductility and energy absorption.
Tests indicate that reinforcing that has actual tensile strength exceeding
the actual yield strength by at least 25% provides substantial yield region.
Additionally, the actual yield strength should not exceed the design yield
strength by more than 18,000 psi. The intent of this limitation is to prevent
the development of undesirable failure modes (such as shear) prior to the
formation of plastic hinges.
6. Reinforcing bar sizes should be smaller than No. 11.
DETAILING
371
7. Straight reinforcing bars should be used to avoid reduced ductility at bends.
8. Continuity with In-plane Members: Every opportunity should be taken to
provide a load path from framing members to plane elements. For example,
beams should be tied to floor slabs, and columns to bearing walls. Particularly in situations when blast forces would be directed against the weak
axis of the member (often the case for beams), this approach can provide
the required blast resistance without having to substantially increase the
dimensions of the member.
9. Balanced Design: Where appropriate, members should be designed to resist the full capacity of supported elements (as defined by a governing
failure mechanism), rather than the design loads transmitted by those elements. For example, it may be desirable for a girder supporting a floor
slab to be designed to resist in-plane forces developed by catenary action
in the slab, even though these forces might be significantly greater than
forces developed from the design loads alone with the slab responding in a
flexural mode.
10. Weak-Beam, Strong-Column: Beam failure is generally preferable to column failure, since the beam failure typically only affects the structure in
close proximity, whereas the column failure could lead to progressive collapse.
11. ACI 318-08 includes provisions for spiral reinforcement (referred to as
“laced” in UFC 3-340-02). Spiral reinforcement provides confinement and
torsion and shear capacity that are superior to stirrups and rectangular hoop
reinforcement. References to selected ACI 318-08 guidelines for spiral reinforcement are included below. UFC 3-340-02 includes more detailed
information on detailing members with laced reinforcement. In general,
the higher performance of spiral reinforcement is most necessary for very
close-in blasts (Z < 1.0), which are beyond the scope of this guide. But
any situation requiring a high level of structural performance merits consideration of spiral reinforcement.
14.3.2 Splices
Splicing can be achieved by lapping bars (No. 11 bars and smaller), mechanical
connectors, or welding. It is good practice to locate splices in noncritical regions,
where low stress states are expected under blast conditions. In particular, splices
should be avoided in locations where plastic hinges are likely to form and at
member intersections. The percentage of bars spliced in a single location should
be limited to 25 percent if possible. Splices should also be separated along the
axial length of a member by at least twice the member depth (Figure 14.1).
Lap splices should satisfy ACI 318-08 Class B requirements, which provide
for a 30% increase in development length over Class A lap splices. Lap splices
should be enclosed by lateral reinforcement conforming to Sections 21.6.4.2,
21.6.4.3, and 21.6.4.5 of ACI 318-08, to provide adequate confinement of the
372
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
Abfy
Abfy
1.3 ld
1.3 ld
1.3 ld
(min)
1.3 ld
2 d (min)
1.3 ld
Figure 14.1 Schematic of ACI 318-08 Lap Splice Spacing and Grouping Recommendations.
concrete surrounding the splices, and to protect the splices from the effects of
spalling. Noncontact lap splices are not recommended. According to UFC 3340-02, Section 4-21, splices of adjacent bars should be staggered by at least the
required lap length, since overlap of splices of adjacent bars is undesirable. Section 21.5.2.3 of ACI 318-08 advises that lap splices in flexural members should
not be used within joints (a) within a distance of twice the member depth from
the face of a joint, or (b) at regions of flexural yielding. Since hinge locations
can be difficult to anticipate for blast loading, welded and mechanical joints are
preferred for blast-resistant design.
If mechanical splices are used, they should satisfy ACI 318-08 Type 2 requirements, which call for at least 125% of the specified yield strength of the bar. UFC
3-340-02 necessitates testing to exhibit the adequacy of mechanical splices under dynamic conditions prior to acceptance for use in hardened structures. UFC
3-340-02 requires that mechanical splices develop both equivalent tensile capacity and ductility of a continuous bar; however, no mechanical splices have been
shown to satisfy this ductility requirement. Thus, where ductility is of paramount
importance, mechanical splices should be avoided.
UFC 3-340-02 suggests that welded splices should be avoided, since welding
may result in a reduction of the ultimate strength and ductility of the reinforcing
steel. However, if welding is necessary, strict adherence to the requirements
of ANSI/AWS D1.4, Structural Welding Code—Reinforcing Steel (American
Welding Society 2005), should permit proper development of the reinforcing
steel strength. Use of testing to demonstrate the performance of welded
reinforcing steel is also good practice.
14.3.3 Columns
Columns subjected directly to lateral air-blast pressures are often the most critical single design consideration. These columns experience large shear forces,
DETAILING
373
out-of-plane deformations that can lead to instability, and reflected pressures that
may have the potential to shatter concrete.
The flexural response of the columns may develop a three-hinged mechanism
with plastic hinge zones located at end supports and at an intermediate point
along the length of the column. The location of the intermediate plastic hinge
can be determined by analysis and depends on the location of the blast source as
well as the height and properties of the column. This analysis can be challenging,
and may require assumptions that render results of little use. For most situations,
the best approach is either to conservatively design the column for the potential
of a hinge anywhere along the entire length, or to make general assumptions of
intermediate hinge location based on the anticipated blast loading. If the column
is in the far field, the blast pressure will be essentially uniform, and the intermediate hinge will form near the center of the column (assuming no other structural
details that would alter the symmetry of the column response). If the column is
in the near field, and subjected to blast from a package bomb, the blast pressures
will be concentrated at the base of the column, and move the intermediate hinge
downward.
The P-Delta moment generated by the inelastic response must be addressed.
Of critical concern is whether the column can develop inelastic response and
maintain sufficient moment capacity to resist the associated P-Delta effect.
FEMA-350, Recommended Seismic Design Criteria for New Steel MomentFrame Buildings (Federal Emergency Management Agency 2000), addresses
this concern for structural steel seismic special moment frames by requiring
the moment capacity at 0.03 radians of inelastic response to be greater than
the nominal plastic moment, Mp . A similar rationale can be adopted for reinforced concrete members, and a synthesis study of available reinforced concrete SMF data indicates that this detailing provides inelastic response of 0.03
radians at the connections. This level of response would equate to the Moderate to Heavy criteria which are typically associated with LOP III and II,
respectively.
Detailing of beam-column joints is discussed in Section 14.3.5 of this chapter,
and per the above discussion it is an essential component of achieving P-Delta
resistance by providing lateral restraint at frame points.
P-Delta resistance within the clear height of the column is achieved with
proper reinforcement detailing so that inelastic response in this region does not
unacceptably reduce the flexural resistance of the column. Detailing for columns
should be based on SMF provisions in ACI 318-08, Section 21.6. The following
subsections present the primary issues to address.
Longitudinal Reinforcement Section 21.6.3.1 of ACI 318-08 limits the area of
longitudinal reinforcement to between 1% and 6% of the gross area. The lower
limit provides that the column will achieve the yield moment before the cracking
moment. The upper limit addresses congestion and avoiding development of high
shear stresses. To minimize the number of iterations, it is typically best to begin
the design process with reinforcing steel at the 6% limit.
374
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
Transverse Reinforcement (Diagonal Shear and Confinement) Transverse reinforcement serves the equally important purposes of resisting diagonal tension
and providing confinement. Confinement from closely spaced hoops protects the
core concrete from being shattered by high, directly applied blast pressures, or
crushed by large deflections and stresses at locations of large inelastic flexural
response. Maintaining the integrity of the core concrete is critical for lateral restraint of the longitudinal reinforcement and allows the concrete to contribute
to the shear capacity of the section. Hoops and ties for confinement are particularly critical in regions of inelastic response. Per the previous discussion of
intermediate hinge location, unless that position can be reliably predicted, it is
recommended that transverse reinforcement be provided per Section 21.6.4 of
ACI 318-08 over the clear height of the column, rather than only over a distance
l0 , as defined in Section 21.6.4.1. Where steel jacketing is provided, transverse
reinforcement can be reduced, since the steel jacket will protect against spalling
and provide the necessary confinement.
Transverse reinforcement spaced to satisfy confinement requirements (Section 21.6.4) will likely satisfy shear requirements for diagonal tension; however,
the requirements of Section 11.4 should always be checked, particularly when
the confinement requirements of Section 21.6.4 are not provided over the clear
height of the column.
Direct Shear Columns must be designed to resist the effects of direct shear as
well as flexure and diagonal tension, although in most cases the direct shear will
not control.
Resistance to direct shear from blast loading can be treated in a manner similar
to that of Section 11.6 of ACI 318-08, which provides minimum requirements for
shear friction reinforcement. The shear friction concept presumes a crack at the
critical shear plane, such that sufficient reinforcing steel is required to maintain
interlock of the rough, mating surfaces. Since the function of the shear friction
reinforcement is to resist the tendency of the shear plane to open, it is the tensile
capacity (not shear) that is of interest; as long as there is sufficient steel, and
it does not yield, the mating surfaces will remain engaged and resist the shear.
Presuming sufficient shear friction reinforcement (as defined ACI 319, Section
11.6), the direct shear capacity of the concrete and any inclined bars that cross
the shear plane can be calculated as
Vn = 0.18 f cd bd + As f yd cos α
Where, f cd
is the product of the concrete compressive strength, DIF and SIF; As
is the area of inclined steel bars; f yd is the product of the steel yield stress, DIF
and SIF; and α is the angle formed by the plane of the inclined reinforcement and
the surface of the shear plane. Note also that since the direct shear phenomenon
occurs very early during the blast application, and before the development of
flexural response, Type I cross section is used. Type I cross section is defined as
the uncracked cross section; that is, the complete concrete cross section can be
counted for resisting the direct shear.
DETAILING
375
7
6
5
KEY
3
4
3
1. Steel jacket – continuous below grade
2. No splice within 2d of steel jacket
3. Splice zone – expected linear behavior
(Type B lap or type 2 mechanical)
4. Hinge zone – no splice
5. No splice within 2d joint
6. Transverse reinforcement continued through joint
7. Transverse reinforcement spaced per ACI 318-08
21.6.4 over entire length (or use round, spiral
reinforced column)
2
1
Figure 14.2 Schematic of Column Detailing Recommendations
Splice Locations Even though hinge location can be difficult to estimate, as
discussed previously, the attempt should be made so that splices can be avoided
at those locations. Splices should never be located in a region that does not have
transverse reinforcement placed per Section 21.6.4 of ACI 318-08.
Additionally, splices should not be located within a joint or within a distance
of twice the member depth from the face of a joint.
Unbraced Length Designing the column to support some or all of the design
load after loss of intermediate lateral support is an indirect means of providing
substantial buckling resistance. This equates to a scenario of the loss of bracing from floor framing, and essentially requires designing the column for an
unbraced length equal to two story heights.
This is not a blast-resistant design method per se, although it clearly relates
to the way the building will resist progressive collapse. However, this is such
an important design consideration that it should be carefully considered, even if
formal progressive collapse design is not utilized. And, this is an essential design
concept whenever a strong-column, weak-beam approach is used.
14.3.4 Beams
This section addresses beams with clear spans greater than four times the beam
effective depth. Beams shorter than this should be treated as deep beams with
shear capacity determined by methods in other sources. Use of shear beams for
376
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
blast-resistant design should be carefully considered due to the potentially brittle
failure modes associated with them.
Exterior beams subjected directly to lateral air-blast pressures experience
large shear forces, out-of-plane deformations that can lead to the development
of plastic hinges, and reflected pressures that may have the potential to shatter
concrete. Interior beams may be subjected to either direct blast pressures in the
form of uplift (for situations in which the interior of the building becomes pressurized) and forces transmitted from other structural members.
Plastic hinge development for exterior beams is similar to that for columns,
as previously discussed. Plastic hinge zones can develop at fixed supports and at
intermediate points along the length of a beam. The exact location of an intermediate plastic hinge can be determined by analysis and depends on the location
of the blast source as well as the span and properties of the column. In lieu of
such complex analysis, two general cases can be considered to bound the problem. Beams subjected to far-field blast load will most likely develop hinges at
supports and at midspan. This assumes that the load applied to the beam from
a far-field blast will be nearly uniform over the column length. Sources of exceptions to this assumption could include very long span beams and beams with
intermediate framing. Beams subjected to near-field blast load will most likely
develop the middle plastic hinge at a location biased to the position of the blast
source.
Simply supported beams can be used in blast-resistant structures if they can be
shown to satisfy the required level of performance. However, the designer must
be aware that the simple supports limit the available potential energy dissipation
and additional margin against unanticipated blast effects.
Detailing for beams should be based on SMF provisions in ACI 318-08, Section 21.5. The following subsections present primary issues to address during the
design process.
Longitudinal Reinforcement Longitudinal reinforcement should comply with
ACI 318-08, Sections 21.5.2.1 and 21.5.2.2. Beams subjected to uplift should
have reinforcement designed to resist both the uplift and rebound.
Transverse Reinforcement (Diagonal Shear and Confinement) Transverse reinforcement serves the equally important purposes of resisting diagonal tension
and providing confinement. Confinement from closely spaced hoops protects the
core concrete from being shattered by high, directly applied blast pressures, or
crushed by large deflections and stresses at locations of large inelastic flexural
response. Maintaining the integrity of the core concrete is critical for lateral restraint of the longitudinal reinforcement, and allows the concrete to contribute
to the shear capacity of the section. Hoops and ties for confinement are particularly critical in regions of inelastic response. Transverse reinforcement should be
spaced to satisfy 21.5.3.2 and provided over the clear span of the beam, unless
specific analysis of yield zone location supports alternate placement. Transverse
reinforcement spaced to satisfy 21.5.3.2 will likely satisfy shear requirements
DETAILING
377
for diagonal tension; however, the requirements of Section 11.4 should always
be checked, particularly when the confinement requirements of Section 21.5.3.2
are not provided over the clear span of the column.
Single-piece hoops as shown in Detail A of Figure R21.5.3 of ACI 318-08 are
preferred to hoops built up from multiple members.
Direct Shear The discussion and recommendations for direct shear in columns
can be directly applied to beams.
Splice Location Even though hinge location can be difficult to estimate, as
discussed previously, the attempt should be made so that splices can be avoided
at those locations. Splices should never be located in a region that does not have
transverse reinforcement placed per Section 21.5.4 of ACI 318-08.
14.3.5 Beam-Column Joints
Joint design is critical to providing an engineered response to blast loading. The
development of inelastic response, which typically accounts for the majority of
energy dissipation for LOP I–III, is directly related to connection performance.
The performance intent of LOP IV is essentially elastic response to the design
blast load; achieving this intent, and also providing margins against unanticipated
load magnitudes, requires equally careful joint detailing.
The performance expectations of the various levels of protection can potentially be achieved with a simple joint between beams and columns, but the tradeoff of this design approach is potentially massive beams.
The monolithic nature of reinforced concrete can be used to great advantage in
the development of robust beam-column joints. Reinforced concrete joints inherently provide substantial deformation continuity between beams and columns.
To provide the level of performance demanded for blast resistance, measures
must be taken to ensure ductility and to protect the core concrete. These goals
are primarily achieved through detailing considerations for the beam longitudinal reinforcement (ductility) and column transverse, confinement reinforcement
(protection of core concrete).
Longitudinal Reinforcement Wherever possible, beam longitudinal reinforcement should be continuous through joints. Splices should not occur within joints,
or within a distance of twice the beam depth from the face of the joint.
Reinforcement that terminates at a joint should comply with Section 21.7.2.2,
which requires the reinforcement to be anchored at the far face of the confined
column core.
Transverse Reinforcement Column transverse reinforcement should be continued through the joint, as described in Section 21.7.3. This section provides
consideration of the similar function that is provided by continuous beam longitudinal reinforcement for joints that have beams framed to all four faces. The
378
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
6
5
4
1
2
3
2
1
KEY
1. No splice w/in 2d from column face
2. Splice zone – expected linear behavior
(Type B lap or type 2 mechanical)
3. Hinge zone – no splice
4. Standard hook for bottom and top reinforcement
(to resist uplift)
5. Longitudinal reinforcement continuous through joint
6. Transverse reinforcement spaced per ACI 318-08
21.5.3 over entire span
Figure 14.3 Schematic of Beam Detailing Recommendations
important consideration is to have confining reinforcement around the column
longitudinal reinforcement within the joint.
Shear Strength Shear strength should be computed per Section 21.7.4. Also of
importance in this section is the definition that a beam is considered to provide
confinement to the joint if a least three-quarters of the face of the joint is covered by the beam. This definition applies to the calculation of shear resistance.
However, relative to the consideration of transverse reinforcement discussed previously, the most important consideration is enclosure of the column longitudinal
reinforcement that continues through the joint. Thus, a beam that satisfies 21.7.4
for shear capacity of the joint should not necessarily negate the use of transverse
reinforcement, spaced as previously discussed, to provide additional protection
of the joint core.
14.3.6 Slabs
The designer must first determine the structural behavior anticipated for the slab.
Slabs can serve numerous roles, including:
Flexural element: This is a conventional role for a building slab. With respect
to blast resistance, a slab will respond as a flexural element to directly applied
blast forces. This can occur with explosions initiated within the building, or
when the interior becomes pressurized by failure of exterior elements. Blast
pressures acting on the soffit of the slab will generate uplift and stresses opposed
to gravity loads.
DETAILING
379
Diaphragm: Slabs that are intended to act as diaphragms for lateral response
to conventional loads can also be used to efficiently distribute blast forces
through the structure, and thus dissipate the blast effects by engaging additional
structural elements and building mass.
Local Restraint: Slabs that are not explicitly intended to perform as a diaphragm can still be used to provide local lateral restraint to beams that are
directly exposed to transverse blast forces. This can be a very effective way to
support beams against the weak-axis bending.
Membrane: Slabs that develop extreme deformations from large, directly applied blast loading may be designed to transition from a flexural response to
a tension membrane response. In addition to careful detailing of the slab, this
mode of response requires careful consideration of the large in-plane forces the
slab will transmit to the surrounding structural elements.
The following subsections are concerned with slab-detailing considerations.
Flexural Reinforcement The flexural reinforcement ratio should not be less
than 0.0018. This is the limit for temperature and shrinkage, when using 60 ksi
reinforcement, given in Section 7.12. In practice, the reinforcement ratio will
almost always exceed this.
Flexural reinforcement should be provided at top and bottom throughout the
slab. For blast scenarios in which uplift is expected, the top reinforcement should
be explicitly designed to resist the uplift forces, and the bottom reinforcement to
resist rebound. In the event that design blast loads will not subject the slab to
uplift, but the slab is part of the structural system that will resist blast effects,
it is good practice to provide approximately equal amounts of top and bottom
reinforcement. This approach will provide reserve capacity for events such as an
unanticipated reversal of curvature or loss of a support.
Bottom reinforcement should be continuous across interior supports. This is
particularly important in the column strip of two-way flat-slab floor systems.
At exterior supports, the reinforcement should be fully developed. This bottom
reinforcement detailing is a precaution against the loss of a support, providing
the structural potential for the slab to bridge across the lost support.
Connected Elements Structural elements connected to a slab should generally
be designed to establish the full capacity of the slab (i.e., balanced design). This
must be carefully considered where the slab is intended to develop tension membrane behavior, because the in-plane forces that will need to be equilibrated by
the surrounding structure will be many times larger than the flexural forces transferred from the slab.
Slabs that serve as lateral restraint for beams directly subjected to blast pressures must have sufficient connectivity for the beam to engage the mass and
in-plane stiffness of the slab.
Fully distributed force transfer between the slab and connected elements is
the preferred mechanism for all situations. For structures with concrete framing
members, this is readily achieved. With steel-framed structures, the best method
380
DESIGN CONCEPTS AND MEMBER DETAILING: CONCRETE
is to cast the member into the concrete slab. For example, for a steel girder supporting a concrete deck, the top flange of the girder should be cast into the deck,
if possible, rather than using shear studs.
14.3.7 Walls
As with slabs, the designer must first determine the structural behavior anticipated for the wall. Walls can serve numerous roles, including:
Gravity System: As part of the gravity load system, walls may experience
uplift from internal pressurization, or, in rare instances, increased compressive
loads from internal overpressures at higher floors.
Flexural Element: A wall will have to be able to respond in flexure in the event
of blast forces applied directly to one face, or unbalanced blast forces applied to
both faces. This can occur with explosions that are set off within a building, or if
the interior becomes pressurized from failure of exterior elements.
Diaphragm: Walls that are intended to act as diaphragms for lateral response
to conventional loads can also be used to efficiently distribute blast forces
through the structure, and thus dissipate the blast effects by engaging additional
structural elements and building mass.
Local Restraint: Walls that are not explicitly intended to perform as a diaphragm can still be used to provide local lateral restraint to columns that are
directly exposed to transverse blast forces.
The following subsections are concerned with wall-detailing considerations.
Reinforcement The minimum web reinforcement ratio should be 0.0025,
which will satisfy requirements for both in- and out-of-plane loads. Reinforcement should be placed in at least two curtains and fully developed at the wall
boundaries. Section 21.9 of ACI 318-08, “Special Structural Walls,” contains
additional useful design recommendations and commentary.
Connected Elements Walls that may be subjected directly to transverse blast
pressure and that also provide gravity support for elements in higher stories
should be constructed with columns at each end and a beam spanning between
the columns. The purpose of this is to provide a gravity load path in the event
that the wall is compromised by the transverse blast load. The beams, columns,
and joints should be designed per previous sections of this chapter.
REFERENCES
This chapter is adapted in large part from information in the PCA’s Blast Resistant Design
Guide for Reinforced Concrete Structures (Portland Cement Association 2009), with
the permission of the Portland Cement Association.
American Concrete Institute. 2008. Building Code Requirements for Structural Concrete
(ACI 318-08) and Commentary (ACI 318R-08). Farmington Hills, MI: American
Concrete Institute.
REFERENCES
381
American Welding Society. 2005. Structural Welding Code—Reinforcing Steel
(ANSI/AWS D1.4). Miami, FL: American Welding Society.
ASTM. 2006. Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement (A706). West Conshohocken, PA: ASTM International.
. 2008. Specification for Deformed and Plain Billet-Steel Bars for Concrete Reinforcement (A615). West Conshohocken, PA: ASTM International.
Bilow, D. and M. Kamara. 2007. Capacity of joints to resist impact loads in concrete
moment-resisting frame buildings. Presented at the Structural Engineering Institute
Structural Congress, Long Beach, CA, May 16–19, 2007.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5-1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2000. Recommended Seismic Design Criteria for New Steel Moment-Frame Buildings (FEMA 350). Washington, DC: Federal
Emergency Management Agency, Department of Homeland Security.
Hayes, J. et al. 2005. Can strengthening for earthquake improve blast and progressive
collapse resistance? Journal of Structural Engineering 131 (8): 1157–1177.
Portland Cement Association. 2009. Blast Resistant Design Guide for Reinforced Concrete Structures. Skokie, IL: Portland Cement Association.
Unified Facilities Criteria Program. 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
15
Blast-Resistant Design Concepts
and Member Detailing: Steel
Charles Carter
15.1 GENERAL
This chapter provides guidance on the design of steel structures for blast loads.
It is adapted from information in two references: “Defensive Design” (Shipe
and Carter 2003) and Defensive Design of Structural Steel Buildings (Gilsanz
et al. 2009), with the permission of the American Institute of Steel Construction
(AISC).
Considerations for defensive design usually fall into one of three general
categories: typical building designs, prescriptive building designs, and
performance-based building designs. Some designs combine prescriptive and
performance-based approaches.
15.1.1 Typical Building Designs
The majority of buildings receive no special treatment other than judicious attention to redundant configurations and robust connection designs so that the structural (and nonstructural) components are tied together effectively. The typical
details used in steel buildings inherently provide for redundancy and robustness,
with the capability for load redistribution through alternative load paths. This
fact has been demonstrated repeatedly when steel buildings have been subjected
to abnormal loadings.
In such cases, the following ideas can be beneficial:
r Configure the building’s lateral systems to provide multiple load paths from
the roof to the foundations. Multiple lines of lateral framing distributed
throughout the building are generally better than fewer isolated systems.
r Provide horizontal floor and roof diaphragms to tie the gravity and lateral
framing systems together.
r Minimize framing irregularities in both horizontal and vertical framing when possible. Horizontal and vertical offsets with copes and/or
383
384
r
r
r
r
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
eccentricities can reduce the available strength at member ends—or require
extensive reinforcement to maintain that strength.
Use multiples of the same shape, rather than changing girder and column
sizes. The additional strength in girders and columns that are heavier for
convenience could cost less or be free. The use of a smaller number of
different shapes in the building means a labor savings in fabrication and
erection, and often more than offsets the cost of the additional steel weight,
which is only around 25% of the cost of the structure once it is fabricated
and erected.
Remember that serviceability limit states indirectly add significant structural redundancy to steel framing. Usually, beams and girders are sized for
deflection or floor-vibration criteria, and girders and columns are commonly
sized for drift control. As a result, these elements have significant reserve
strength.
Use typical shear, moment, and/or bracing connections judiciously. Reserve
strength is gained at low cost if connection details are clean. It costs little
to fill the web of a girder with bolts using a single-plate or double-angle
connection.
Recognize other sources of redundancy and robustness inherent in steel
buildings, including: the common overstrength in the steel materials and
connecting elements, membrane action in the floor and roof diaphragms,
and the strength and stiffness contributions of nonstructural components.
With little—and sometimes no—modification, steel framing provides redundancy and robustness.
15.1.2 Prescriptive Building Designs
Some buildings receive special treatment through the application of prescriptive
criteria for design that go beyond those in the basic building code. While there are
a variety of prescribed criteria, such as the removal of a single column, there is
usually no attempt to characterize the exact effects of the blast. Instead, the goal
could be to reduce the probability of progressive collapse in areas not directly
affected by the blast.
In such cases, several strategies can be employed:
r Use a perimeter moment frame. This can result in a system that is significantly robust. As an extreme example, the perimeter moment frame
in the tube structure of each of the World Trade Center towers spanned
a hole about 140 ft wide before succumbing to the combined effects of
structural damage and fire, each likely at their lifetime maximum values.
More likely, the prescribed criterion will be the removal of a single column. In some cases, the framing will have enough redundancy to accept
column removal without modification. If not, the column spacing can be
GENERAL
385
reduced or the framing hardened by increasing size or switching to
composite construction.
r Use a strong story or floor. This solution can be a truss system with diagonals or a Vierendeel truss system, incorporated into a single story or multiple stories in the building. Additionally, a single, strong floor with heavier
framing and moment connections throughout could carry or hang at least
10 floors. The World Trade Center towers also demonstrate this solution.
The damaged core columns in each tower to some extent hung from the
hat trusses, which created strong-story framing at the top of each tower.
The link between the core gravity columns and the perimeter tube framing carried more than 10 stories in one tower and more than 25 stories in
the other. Nonetheless, exercise caution when considering hat-truss framing to create a strong story. Unless specifically designed for progressivecollapse resistance, hat trusses normally reduce the level of reserve strength
and redundancy because of the efficiency they allow in the structural
system.
r Consider other innovative solutions. One particularly innovative solution
was used in the U.S. Federal Courthouse in Seattle, WA. The building
has a steel-framed composite core and gravity steel framing around it. The
perimeter has steel cables banding it to prevent progressive collapse should
a column be lost. The result is one of the most open and inviting blasthardened buildings built to date.
There are other potential prescriptive criteria, but structural solutions usually
flow from criteria like those described above. Most often, there is no attempt
to rigorously assess the actual blast effects and the emphasis is on arresting the
effects of the blast.
15.1.3 Performance-Based Building Designs
Performance-based criteria are used for a small number of buildings, normally
ones that are government-related, high-risk, or high-profile. Performance criteria vary, but generally require that the building withstand the effects of the blast,
protect the occupants of the building, and/or maintain a defined level of operability. The nature and characteristics of the threats are identified realistically and
modeled in the design. Note that the performance criteria affect more than the
structural frame, and often require nonstructural elements, such as blast-resistant
windows, special site layouts, and site perimeter protection.
Performance-based criteria are used for a comparatively small number of
buildings, where the nature and characteristics of the threat are realistically identified and modeled in the design. When the threat, building characteristics, and
performance criteria are known, the solution is a matter of design for the dynamic loading of the blast, including mitigation of damage and the associated
progressive collapse potential. The following sections illustrate this approach.
386
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
15.2 BLAST EFFECTS ON STRUCTURAL STEEL AND COMPOSITE
STRUCTURES
Steel’s response to the characteristics of a blast affects the loading for several
reasons, discussed in the following subsections.
15.2.1 Member Ductility
Structural steel has a linear stress-strain relationship up to the yield stress, but
then can undergo extreme elongation without an increase in stress—about 10 to
15 times the amount needed to reach yield. Stress then increases in the “strain
hardening” range until a total elongation of about 20 percent to 30 percent is
reached. This response has benefits beyond routine design-level forces for resisting the effects of a blast. The ductility ratio m, defined as the maximum deflection
over the elastic deflection, commonly is used to account for this effect.
15.2.2 Connection Ductility
Steel connections can be extremely ductile. As in seismic design practice, connections can be configured so that yielding limit states will control over fracture
limit states. However, there is still significant ductility in many limit states that
involve fracture. Block shear rupture, for example, is a fracture limit-state accompanied by significant deformation prior to the actual failure.
Additionally, some limit states have “failures” that, in blast applications, can
be considered benign or even beneficial. For example, there is little structural
consequence to connection slip or bearing deformations at bolt holes due to blast
effects. Moreover, these phenomena consume energy when they occur, progressively reducing the blast effect through the structure.
15.2.3 Overstrength
Structural steel usually is stronger than the specified minimum strength. For
example, ASTM A992, Standard Specification for Structural Steel Shapes
(ASTM 2006) has a specified minimum yield strength Fy = 50 ksi. From the
American Institute of Steel Construction (AISC) Seismic Provisions for Structural Steel Buildings (American Institute of Steel Construction 2005a), it has an
expected yield strength Fye of 55 ksi. This supports the practice endorsed by the
Department of Defense Explosives Safety Board, which allows the use of an average yield strength increase factor (generally a 10 percent increase in the yield
strength) for blast design.
15.2.4 Beneficial Strain-Rate Effects
Structural steel benefits from an increase in apparent strength when the rate of
loading is rapid. The yield point increases substantially by a factor that is called
BLAST EFFECTS ON STRUCTURAL STEEL AND COMPOSITE STRUCTURES
387
the Dynamic Increase Factor for yield stress. The ultimate tensile stress also
increases, but not as greatly as the yield stress. The total elongation at failure
typically remains unchanged or decreases slightly because of the increased strain
rate. The modulus of elasticity is unaffected by the strain rate.
15.2.5 Beneficial Effects of Composite Construction
The use of composite construction can have benefits for blast-design applications
due to the mass effect of composite systems with steel elements (490 lb/ft3 )
and concrete elements (150 lb/ft3 ). The inelastic action in a composite system
generally will limit deflections and local deformations, and partially mitigate
rebound effects through the damping effect of concrete cracking.
15.2.6 Perimeter Column Design
At the building perimeter, the columns will be loaded by the blast through the
facade. A critical design decision is the selection of an appropriate tributary area
for the column, which must be based upon the expected performance of the facade in the blast. Will the facade and its connection to the structure be such that
the full blast load will be delivered to the framing, or will the facade components
be shattered in the blast before the load can be delivered through them?
The tributary area is then used with the maximum total dynamic pressure and
the dynamic load factor to find an equivalent static load on the column. Either a
rigorous beam-column design approach or the simplified approach suggested in
UFC 3-340-02 (Unified Facilities Criteria Program 2008) can be used:
r Support conditions for the column are taken as if the column were a simplespan beam.
r The moment perpendicular to the plane of the facade is taken as wl2 /8 based
upon the blast loading.
r The moment in the plane of the facade is a function of the magnitude of the
strong-axis moment and the angle of incidence of the blast to the column.
r The axial load applied concurrently with the blast load is based upon the
full dead load and one-quarter of the live load.
Based upon these loadings, UFC 3-340-02 recommends checking limit states
like flange-local buckling, web-local buckling, shear yield, and interaction between flexural and axial forces.
15.2.7 Perimeter Girder Design
The consequences of a girder failure normally are not as high as the consequences of a column failure. When designing the girder to the same performance
level as the columns, use the foregoing approach to column design for the girder
388
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
design, without the axial load. Otherwise, a less restrictive approach can be used,
permitting inelastic deformations. It is important to consider the differing support conditions between the top and bottom flanges of the girder. Often this can
be mitigated by orienting the infill beams to provide restraint to the girder bottom flange along the span. Also, the girder end connections often will be best
configured as moment connections.
15.2.8 Slab Design
The blast pressure could subject the slab to significant uplift, depending upon
the expected performance of the facade in the blast. As with girder design, the
consequences of a local slab failure normally are not as high as the consequences
of a column failure, so it could be sufficient to utilize the typical reinforcing steel
in the upper portion of the slab. If a higher performance objective is established,
the blast pressure can be applied to the slab and the reinforcement selected to be
appropriate for the loading.
15.3 ANALYSIS AND DESIGN OF STRUCTURAL MEMBERS
Failure modes determine the mechanism of collapse for the structural element.
Deformation criteria classify the response of the element. The primary failure
modes used to design structural elements are breaching, tension-compression,
bending, shear, and axial-bending interaction.
Member failure is defined through support rotation and ductility ratios for use
with single-degree-of-freedom (SDOF) and simplified multi-degree-of-freedom
(MDOF) systems. Strain-based failure criteria may also be justified if strains are
calculated using finite element methods that characterize the material properties, the details of construction, and many of the possible failure mechanisms.
Ductility is the ratio between the maximum deflection and the maximum elastic
deflection. This parameter is smaller than one (1) if the behavior is elastic, and
larger than one (1) if the behavior is plastic.
Energy dissipated along the load path is ignored, and it is assumed that the
elements are absorbing all of the energy, which is a conservative assumption.
The designer should give attention to the reactions and the connections along the
load path from one member to another one, as these will be required to transfer
the full energy of the system.
15.4 STEEL MATERIAL PROPERTIES FOR BLAST DESIGN
The properties used in conventional design of steel buildings are modified when
considering blast loadings to account for actual strength level and dynamic
effects.
STEEL MATERIAL PROPERTIES FOR BLAST DESIGN
389
15.4.1 Strength Increase Factor (SIF)
For steel grades with Fy = 50 ksi or less, the average yield stress in current production is approximately 10% larger than this minimum specified value. Therefore, for blast design the yield stresses should be multiplied by a Strength Increase Factor (SIF) of 1.10. For higher-strength grades, this average is smaller
than 5% larger; therefore, no factor is used on those grades. The tensile strength
Fu is not factored in any case.
15.4.2 Dynamic Increase Factor (DIF)
Steel mechanical properties vary with the time rate of strain. As compared with
the static values normally used in design, the following properties vary for dynamic loading:
r The yield point increases substantially.
r The ultimate tensile strength increases slightly.
r Modulus of elasticity does not vary, and the elongation at rupture either
remains constant or is slightly reduced.
The factor used to modify the static stress due to dynamic load is the Dynamic Increase Factor (DIF). These factors are defined in Design of Blast Resistant Buildings in Petrochemicals Facilities (American Society of Civil Engineers
1997) and UFC 3-340-02.
The values summarized in Table 15.1 are based upon an average strain ratio of
0.10 in./in./sec, which is characteristic of low-pressure explosions. Higher values
of strain ratio will result in larger values of DIF. UFC 3-340-02 provides values
of DIF for different average strain rates.
Common grades in use today, such as ASTM A992, A500, and A53 (ASTM
2006, ASTM 2007b, ASTM 2007a) are not addressed in the aforementioned
Table 15.1 Dynamic Increase Factors (DIF) for Structural Steel
DIF (Low Pressure)
Yield Stress
Material
Bending/Shear
Tension/Compression
Ultimate Stress
A36
1.29
1.19
1.10
A588
1.19
1.12
1.05
A514
1.09
1.05
1.00
A446
1.10
1.10
1.00
390
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
references. All are normal construction grades, similar to ASTM A36 (ASTM
2008), and suitable for use with the same DIF values.
15.4.3 Dynamic Design Stress
Based on the expected ductility ratio and/or the damage allowed in the structural
element, different dynamic yield points are defined by UFC 3-340-02. If the
ductility ratio is smaller than or equal to 10, the dynamic design stress is:
f ds = f dy = SIF · DIF · f y
(15.1)
If the ductility ratio is greater than 10, the dynamic design stress is:
f ds = f dy +
f du − f dy
4
(15.2)
where
f du = DIF · f u
(15.3)
The dynamic design stress for shear is
f dv = 0.55 f ds
(15.4)
For typical design, unless governed by more detailed requirements such as
Design of Blast Resistant Buildings in Petrochemical Facilities (ASCE 1997),
UFC 3-340-02, or other established criteria, a simplified value may be used as
follows:
f ds = 1.3 f y
(15.5)
f du = 1.05 f u
(15.6)
15.5 DESIGN CRITERIA FOR BLAST DESIGN
15.5.1 General
Due to the nature of blast loading, plastic design is required, and is measured by
support rotations and ductility. Different codes can be used to define the strength
of the elements, but no safety factor should be used in those calculations. Local
and global stability should be addressed in those elements where plasticity must
be achieved and ductility criteria must be used.
Design and failure criteria are based on the results of explosive tests conducted
by the U.S. government and reported in publications such as the Tri-Service
DESIGN CRITERIA FOR BLAST DESIGN
391
Manual by the Department of Defense, and the Protective Design Manual by
the U.S. Air Force.
The designer should determine, at the beginning of the project, the acceptance
criteria to be used based on the level of protection desired for the building. Most
projects have predefined criteria that must be used.
15.5.2 Load Combinations
In the absence of other governing criteria, the following load combination should
be used:
1.0D + 0.25L + 1.0B
(15.7)
where D is the dead load
L is the live load
B is the blast load
15.5.3 Resistance Factor and Factor of Safety
For Load and Resistance Factor Design (LRFD), the resistance factor φ is not
used in blast design. For Allowable Stress Design (ASD), the safety factor
is not used in blast design. Thus, the strength is simply taken as the nominal
strength, without reduction.
15.5.4 Local Buckling
Blast design is based on the ultimate strength of the elements and ductility of the
system. Structural members subject to blast loads should be capable of undergoing plastic deformation. To allow hinges to form in the elements, sections must
be flexurally compact in accordance with criteria in Chapter B of the AISC Specification for Structural Steel Buildings (American Institute of Steel Construction
2005b), unless the effects of local buckling can be tolerated in the response. The
dynamic design stress should be used in the calculation of the limiting slenderness to establish the local buckling criteria.
15.5.5 Lateral-Torsional Buckling
Flexural members should be braced at an interval that is small enough to permit
plastic hinge formation. Essentially, this removes lateral-torsional buckling as
the controlling limit state in beams.
15.5.6 Deformation Criteria
Support rotation is defined as the tangent angle at the support formed by the
maximum beam deflection. In the plastic range, this value, neglecting the elastic
deformation, can be related to the plastic hinge rotation. Note that if the hinge is
392
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.1 Hinge Rotation versus Connection Rotation
not formed at the center of the beam, the support rotations are different at both
sides, and the maximum rotation should be considered as shown in Figure 15.1.
Ductility, µ, as used in blast design, is defined as the ratio between the maximum deflection (m ) and the elastic deflection E ).
µ=
m
E
(15.8)
ASCE’s Design of Blast Resistant Buildings in Petrochemical Facilities
(American Society of Civil Engineers 1997) classifies the deformation range in
three different damaged stages: low, medium, and high response as a function
of the damage in the building. UFC 3-340-02 also classifies the response as a
function of the protection provided by the structural elements. This section is
based upon a low level of protection that implies a high response as the intent is
to provide a design that avoids imminent collapse.
For different structural elements, this guide follows the response criteria as
shown in Table 15.2. The rotation criteria in Table 15.2 refer to support rotation.
The criteria are defined for the behavior of a single element. These criteria apply
to the design of the members, and the connection between the members may
have different criteria.
Table 15.2 Response Criteria for Structural Steel
for a Low Level of Protection
Element
Response Ratio (Ductility) Rotation
Open-web steel joists1
µ<2
θ < 3◦
Steel beams
µ < 20
θ < 10◦
Steel columns
µ<5
θ < 2◦
1
Response ratio (ductility) controlled by downward loading, and
rotation controlled by upward loading
DESIGN CRITERIA FOR BLAST DESIGN
393
As an example, assume a beam with a 30-ft span that starts to yield at midspan
when the deflection is 4 in. A support rotation of 10◦ implies a deflection of 15 ×
tan(10◦ ) = 31 in. A ductility of 1 implies a deflection of 4 in. Thus, a ductility of
20, again taken from Table 15.2, would require a deflection of 80 in. Obviously,
this level of ductility is not achievable, and the support rotation controls.
Other more restrictive criteria exist, such as Corps of Engineers Program Design and Construction (COE PDC) criteria, and are usually used in commercial
and governmental buildings.
Ductility limits are linked to a given mode of response. A flexural ductility is
different from shear or tension. For steel elements not shown in Table 15.2, other
sources can be consulted, including UFC 3-340-02 and Design of Blast Resistant Buildings in Petrochemicals Facilities. Although it is intended for seismic
design, FEMA 356 (Federal Emergency Management Agency 2000) provides
useful information.
UFC 3-340-02 defines the blast criteria and the structural properties used in
Figure 15.2.
15.5.7 Detailing for Specific Failure Modes:
Failure modes include breaching, tension, compression, shear, flexure, and combined loading.
r Breaching: Blast loads in contact with or in very close proximity to an element. May cause failure before a more typical structural response like flexure can occur, due to the high pressure produced by the explosion. Generally, if the scaled distance is below two, it is possible for the element to
breach before the overall response of the structural element starts.
For very close charges, in addition to the air blast, temperature and the
shock wave are important. An explosion in direct contact with a floor or wall
will interact directly with it and induce a shock wave inside it. The speed
and magnitude of the shock waves can cause the floor or wall to crack internally. For example, when the shock wave reaches the opposite face of a concrete wall, a section of concrete may spall or separate from the wall because
the energy of the shock wave exceeds the tensile strength of the material.
This behavior can be estimated using information in Figure 15.3, from
the Protective Construction Design Manual, ESL TR 87-57 (Air Force Engineering and Services Center 1989), based upon the thickness (T, ft), the
standoff distance (R, ft), and the charge (W, equivalent lbs TNT). Depending on the performance requirements, the slab may be designed to allow a
local breach in the bay closest to the blast, yet withstand the blast in adjacent bays. Typically, the standoff and charge weight are specified, given the
threat analysis. Other breaching curves can be found in ESL-TR-87-57 or
UFC 3-340-01 (Unified Facilities Criteria Program 2002).
r Tension: According to AISC’s Specification for Structural Steel Buildings,
the maximum strength for elements under tension is based on yielding of
394
Figure 15.2 UFC 3-340-02 Criteria
DESIGN CRITERIA FOR BLAST DESIGN
395
Figure 15.3 Protective Construction Manual Breach Chart
the member and is defined by
Pu = φ f y A g
(15.9)
where A g = the gross area of the element
f y = the material yield stress
φ = the resistance factor, 0.90
For blast loading, φ is not used, and f y = f ds is the dynamic design stress
defined in Equations 15.1 and 15.2.
Axial hinge properties for these kinds of elements are elastic-plastic,
based on the ultimate load defined above. If necessary, for computational
stability, a small strain hardening slope, such as 0.1%, can be used in the
plastic area.
r Compression: The strength of compression elements, based on the AISC
Specification for Structural Steel Buildings, is governed by the following
equation:
Pu = φ f cr A g
where A g = the gross area of the element
f cr = the critical stress
φ = 0.90
(15.10)
396
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For blast loading, φ is not used, and f cr = the critical stress by the appropriate AISC equation substituting f ds for f y . For a tension-compression
hinge, the recommendations of FEMA 356 can be used, if no other proper
criterion is defined.
r Shear: For I-shapes it is assumed that moment is carried primarily by the
flanges, while shear is carried primarily by the web. Thus, moment-shear
interaction is neglected. The yield strength of elements in shear, based on
the AISC Specification, is governed by the following equation:
V p = φ f v Aw
(15.11)
where Aw = the area of the web
f v = 0.6 f y
φ = 1.00 for most W-shapes
For blast loading, φ is not used, Aw is the area of the web, and f v = f dv is
the dynamic yield stress; f v = 0.55 f y in UFC 3-340-02 (Unified Facilities
Criteria Program 2008). Note that connection strength is, in most of the
cases, more critical than the shear strength of the beam.
r Flexure: Based on plastic response allowed for blast design, a plastic hinge
will be allowed to form in the structural element. The assumption that the
plastic hinge is concentrated at a section will be taken as adequate for practical purposes, even though deflection values may not be accurate.
The definition of plastic moment is a function of the ductility ratio expected in the structural element. Based on Figure 15.7 of UFC 3-340-02, if
the ductility ratio is smaller than 3, the full plastic moment cannot be developed by the section. Thus, an average between the elastic and plastic section
modulus is used:
M p = f ds
S+Z
2
(15.12)
If the ductility ratio expected is larger than 3, the full plastic moment can
be developed. Thus,
M p = f ds Z
(15.13)
For hand calculations, the behavior of the section is assumed elasticperfectly plastic. For hand calculations in this guide, the plastic moment
strength of the section is given by Equation 15.13. To avoid computational instabilities, the hinge properties used for computer calculations
should include plastic hardening, where the yield moment is defined by
Equation 15.12 and the ultimate moment is defined by Equation 15.13 as
EXAMPLES
397
recommended in UFC 3-340-02. A criterion for lateral bracing is also given
in UFC 3-340-02 although not discussed here.
r Combined Forces: Provisions for combined axial and bending effects are
presented in the AISC Specification. The strengths and resistance factors
defined above are used in the appropriate interaction equations.
15.6 EXAMPLES
15.6.1 Example 1—Determining Capacities
The strength of the elements determined in these examples is based on UFC
3-340-02. For element design, it is assumed that there is no energy dissipation
along the load path. A more accurate and less conservative procedure is to use
the dynamic reactions of one member on another along the load path.
When using member reactions to load following members, natural frequencies
should be compared. If the natural frequencies are close, a simultaneous solution
is required to account for the interaction between the two members. If the period
of the primary element (i.e., beam) is at least twice the period of the secondary
element (i.e., girder), they can be treated as individual SDOF structures on unyielding supports. If not, an MDOF solution of the same system should be used.
Material Properties All calculations for this example are done for steel with
the following properties:
r Fy = 50 ksi
r Fu = 70 ksi
For a ductility ratio µ ≤ 10, the dynamic design stress, defined in Equation
15.5, is:
f ds = f dy = 1.3Fy = 1.3 × 50 = 65 ksi
For a ductility ratio µ > 10, the dynamic design stress, defined in Equation
15.2, is:
f du − f dy
1.05 × 70 − 65
= 65 +
= 67 ksi
f ds = f dy +
4
4
where: f du = 1.05fu from Equation 15.6.
The dynamic design shear stress, defined in Equation 15.4, for a ductility ratio
µ ≤ 10 is:
f dv = 0.55 × f ds = 0.55 × 65 = 36 ksi
398
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.4 Tension Hinge Property
Tension-Only Rod Brace A tension rod with 3/4 -in. diameter is used. Because
the rod has upset ends, we can use the full cross section area. The tension strength
is obtained as:
Tmax = A f ds = π ×
3/4
2
2
× 65 = 28.7 kips
The hinge property for this rod follows the elastic-plastic curve shown in
Figure 15.4. A slight slope is included in the plastic region for computational
convergence purposes.
Given a maximum horizontal displacement of 4.2 in., the maximum rotation
at the base of the column for this displacement is:
α = atan
H
4.2
= atan
15 × 12
= 1.34 deg < 2 deg
For a small-displacement formulation, the elongation in the rod can be determined. First, the angle of the diagonal is determined:
θ = atan
H
L
15
= atan
35
= 23.2 deg
EXAMPLES
399
Then, the elongation, assuming the top of the column moves horizontally:
L = cos θ = 4.2 × cos 23.2 = 3.86 in
Thus, the elastic deformation of this rod is:
L el =
f ds L
65 × 38 × 12
=
= 1.02 in
E
29000
Therefore, the ductility demand in the rod is:
µ=
3.86
= 3.78
1.02
Diagonal Brace (in Tension and Compression) An ASTM A500 Grade C
HSS6 × 6 × 3/4 is used. This section has an area of 5.24 in.2 and a radius of
gyration of 2.34 in. The brace length is 17 ft.
The slenderness of this element is:
1 × 17ft × 12in. ft
2π 2 E
KL
=
= 88 < Cc =
= 94
λ=
r
2.34 in
f ds
The buckling stress is obtained from UFC 3-340-02 as:
⎛
2 ⎞
Kl
⎟
⎜
r
⎟
⎜
⎟ f ds
⎜1 −
⎝
2Cc2 ⎠
Fa =
5 3
+
3 8
Kl
r
Cc
−
1
8
Kl 3
r
Cc3
(15.14)
Substituting values, the nominal buckling stress is:
882
1−
× 65
2 × 942
Fa =
= 19.2 ksi
5 3 88 1 883
+ ×
− ×
3 8 94 8 943
The denominator in the above equation (which equals 1.92), is the factor of
safety. Hence, the compression capacity without the factor of safety is:
Pn = 1.92AFa = 1.67 × 5.24 × 19.2 = 193 kips
The tension capacity for this element is:
Pu = A f ds = 5.24 × 65 = 341 kips
400
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.5 Tension-Compression Hinge Properties
Using information from FEMA 356 (Federal Emergency Management
Agency 2000), a tension-compression hinge can be developed, as shown in Figure 15.5. This asymmetric curve includes the buckling ductility. In the tension
zone, the hardening slope is modified to account for the dynamic yield stress
for the yield axial force and the dynamic ultimate stress for the ultimate axial
force. In the compression zone, a hardening slope of 0.1% is introduced to avoid
computational instabilities.
Column Assume the column is a W12×53 with an effective length KL = 15 ft.
The section properties are:
A = 15.6 in.2
rmin = r y = 2.48 in.
For this column:
Cc =
2π 2 E
=
f ds
2 × π 2 × 29000
= 94
65
EXAMPLES
401
The maximum slenderness of this element is:
15 × 12
KL
= 72.6 < 94
=
ry
2.48
The buckling stress is:
72.62
1−
× 65
2 × 942
Fa =
= 24 ksi
5 3 72.6 1 72.63
+ ×
− ×
3 8
94
8
943
And the compression capacity without the factor of safety is:
Pu = 1.90AFa = 1.90 × 15.6 × 24 = 711 kips
The tension capacity of this element is:
Tu = A f ds = 15.6 × 65 = 1014 kips
Beam Assume the beam is a W12×35, 24-ft long, and it has chevron braces
connected to it at midspan. This shape has:
I = 285 in4
S = 45.6 in3
Z = 51.2 in3
It is modeled as a simply supported beam with a concentrated load at midspan,
where the braces meet.
For ductility ratios smaller than 3, the elastic-plastic bending moment is given
by
M pl =
S+Z
45.6 + 51.2 65
f ds =
×
= 262 kip − ft
2
2
12
The elastic deflection for this moment due to the concentrated load at the
midspan is
=
M pl L 2
262 × 12 × (24 × 12)2
=
= 2.63 in.
12EI
12 × 29,000 × 285
This deflection gives an elastic rotation:
θ = arctan
L 2
2.63
= arctan
24 × 12 2
= 1.05 deg
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Moment
402
Mult = Z × fds (µ > 10)
Mpl =
S+Z
fds
2
Rotation
2θult
2θ
Figure 15.6 Beam Hinge Moment-Rotation
Note that the hinge rotation is twice this support rotation. These parameters
define an elastic-perfectly plastic moment-rotation curve. Since many programs
have convergence problems with a perfectly plastic zone, a sloped line should be
introduced in this plastic region. This slope is obtained from Figure 15.6 for ductility greater than 10. Therefore, the ultimate bending moment is computed as:
Mult = Z × f ds (µ > 10) =
51.2 × 67
= 286 kips − ft
12
The maximum rotation from UFC 3-340-02 is:
θult = min (10 deg, 20 × θ) = min (10 deg, 26 deg) = 10 deg
Again, note that hinge rotation is twice this support rotation.
15.6.2 Example 2—Design and Analysis for Blast Loads on Members
Figures 15.7 and 15.8 illustrate the method used in this example for a simply
supported beam with load and mass uniformly distributed.
The same transformation as shown in Figure 15.7 can be done by multiplying
only the mass by the load-mass factor.
MSDOF = M K L M = M
KM
KL
The KLM approach is simpler because it only uses one transformation factor.
This example uses the load factor KL and the mass factor KM because they have
a more physical interpretation.
EXAMPLES
403
EQUIVALENT SDoF PROCEDURE
Continuous System
Discrete System
FSDoF(t)
F(t ) = p(t ) × L
MSDoF
M=m×L
KSDoF
L
Mass = M = m × L
Equivalent Mass = MSDoF = M × KM
Load = F(t ) = p(t ) × L
Equivalent Load = FSDoF (t ) = F(t ) × KL
Elastic Stiffness = K =
384 × E × I
5 × L3
Equivalent Stiffness = KSDoF = K × KL
Plastic Stiffness = Kpl = 0
Yield Force = Ryield =
8 × Mp
L
Equivalent Yield Force = Ryield, SDoF = Ryield × KL
Figure 15.7 MDOF to SDOF Simplification
Wind Girt The wind girt shown in Figure 15.9 has been designed as an MC8
× 20 (ASTM A572 Grade 50; Fy = 50 ksi; Fu = 65 ksi) for wind effects with
a deflection limitation of L/260. The blast deflection criteria given in Table 15.2
show that the ductility should be less than 20, and support rotation should be less
than 10 degrees.
The duration of the blast is 6.2 milliseconds, and the peak load for the 79.5
psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 5 × 25 = 1431 kips
1000
Also given, the load factor K L = 0.50 and the mass factor K M = 0.33.
As a preliminary design, the support rotation criterion is used because it does
not imply the knowledge of the actual section used. Rigid-perfectly plastic behavior is assumed, and the element is sufficiently braced against lateral-torsional
buckling.
404
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
SDoF NON-LINEAR DYNAMIC SOLUTION SYSTEM
System to be solved
FSDoF(t)
MSDoF Structural Period = T
KSDoF
IF
td
< 1
T 10
IMPULSIVE SOLUTION
IMPULSE ENERGY = STRAIN ENERGY
RESULTS
SOLUTION
1 < td
TIME-HISTORY DYNAMIC
IF
10 T
Fpeak, SDoF
DYNAMIC ANALYSIS CAN BE SOLVED:
- INTEGRATION OF MOTION EQUATION
- CHARTS
- SDoF SOFTWARE
FSDoF (t)
td
t
Figure 15.8 SDOF Solution
Figure 15.9 Example Girt
SDoF TO MDoF
- TIME AND DISPLACEMENT
ARE INVARIANT
- FORCE AND REACTION
SHOULD BE TRANSFORMED
EXAMPLES
405
For this rotation criterion, the maximum displacement allowed is:
max =
25 × 12
× sin (10 deg) = 26 in.
2
The self weight of the facade is 40 psf. The total weight of the system is:
w=
40 × 25 × 5
= 5 kips
1000
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33. Therefore, the parameters used in the discrete system are:
Fpeak,SDOF = Fpeak K L = 1431 × 0.5 = 715 kips wSDOF
= wK M = 5 × 0.33 = 1.65 kips
The equivalent impulse in this single-degree-of-freedom system is:
ISDOF =
F peak,SDOF tblast
715 × 6.2 × 10−3
=
= 2.2 kips − sec
2
2
The total energy produced by the blast load in the single-degree-of-freedom
is:
WImpulse,SDOF =
2
ISDOF
=
2m SDOF
2.22
= 566 kips − in
1.65
2×
386
To limit the maximum displacement to comply with the support rotation criteria, the single-degree-of-freedom yield force should be:
R yield,SDOF =
WImpulse,SDOF
566
= 21.8 kips
=
max
26
The yield force for the continuous system is:
R yield =
21.8
= 43.6 kips
0.5
For this yield force, the maximum moment is:
M pl =
R yield L
43.6 × 25
=
= 136 kips − ft
8
8
406
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For a maximum tensile strength of 65 ksi, the plastic modulus should be
greater than:
Z min =
136 × 12
= 25 in3
65
Based on this preliminary design, there is no MC8 strong enough to support
the blast. There are several possible modifications to improve the behavior of the
system: Increase the excited mass, increase the strength-stiffness of the system,
or decrease the blast load by integrating a variable blast pressure that is a function
of the distance to the charge along the girt.
In this example we will increase the steel section to a W8 × 28 (ASTM A992;
Fy = 50 ksi; Fu = 65 ksi):
Z x = 27.2 in3
M pl = Z f ds =
27.2 × 65
= 147 kip − ft
12
Ix = 98 in4
el =
5 147 × 12 × (25 × 12)2
5 M pl L 2
=
= 5.82 in
48 E I
48
29,000 × 98
The strength parameters to use in the dynamic calculation are:
8M pl
8 × 147
=
= 47.0 kips for the SDOF
L
25
= R yield K L = 47.0 × 0.5 = 23.6 kips
R yield =
R yield,SDOF
R yield
47
= 8.08 kips in for the SDOF
=
el
5.82
= KK L = 8.08 × 0.5 = 4.04 kips in
K =
K SDOF
The mass and load do not change from the latest calculations.
The period of the system is:
T = 2π
m SDOF
=2×π ×
K SDOF
1.65
= 0.20 sec
386 × 4.04
The structural period (0.20 sec) is more than 10 times longer than the load
duration (0.0062 sec), hence the assumption of impulsive load is correct.
Figure 15.10 shows the displacement computed using SDOF software. The
maximum deflection is 28.8 in. > 26.0 in.; hence, the rotation criterion is not
met by a slight margin.
EXAMPLES
407
Figure 15.10 SDOF Displacement
The elastic displacement for this beam is 5.82 in.; therefore, the ductility ratio
for this system is:
µ=
28.83
= 5 < 20
5.82
The maximum reaction happens at the time of the maximum response. In an
impulse load, at the time the response is maximum, the load is zero. For a longer
blast duration, the reaction is a combination of the direct load and the capacity
of the resisting element. According to UFC 3-340-02, the maximum shear force
in plastic design can be obtained as the shear reaction that corresponds to the
elastic limit of the section:
V =
R yield
47.2
=
= 23.6 kips
2
2
The dynamic shear capacity is:
V p = f dv Aw = 36 × 2.30 = 82.8 kips > 23.6 kips
where f dv is the blast shear capacity.
Aw is the area of the web
408
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Hence, the section can support the shear. The connection should be designed
for this capacity.
Facade Column The column shown in Figure 15.11 has been designed as a
W12 × 53 (ASTM A992; Fy = 50 ksi; Fu = 65 ksi) for wind effects with a
deflection limitation of L/260. The properties of this section are:
A = 15.6 in.
Ix = 425 in .4
r y = 2.48 in.
Z x = 77.9 in.3
r x = 5.23 in.
As for the wind girt, the duration of the blast is 6.2 milliseconds, and the peak
load for the 79.5 psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 5 × 25 = 1431 kips
1000
Also given, the load factor K L = 0.50 and the mass factor K M = 0.33.
Columns are designed to remain elastic; therefore, the maximum bending capacity should not be reached. As a preliminary design, the column will be designed without axial compression. Combined bending compression formulation
will be used to check the adequacy of this element.
Figure 15.11 Example Column
EXAMPLES
409
As a preliminary and conservative assumption, the girt is assumed infinitely
rigid, and all the blast pressure is absorbed by the column.
The self weight of the facade is 40 psf. The total weight of the system is:
w=
40 × 25 × 15
= 15 kips
1000
The peak load for the 79.5 psi peak blast pressure is:
Fpeak =
79.5 × 12 × 12
× 15 × 25 = 4293 kips
1000
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and the stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33.
Therefore, the parameters used in the discrete system are:
F peak,SDOF = 0.5 × 4293 = 2146 kips
wSDOF = 0.33 × 15 = 4.95 kips
The equivalent impulse in this single-degree-of-freedom system is:
ISDOF =
F peak,SDOF tblast
2146 × 6.2 × 10−3
=
= 6.65 kips − sec
2
2
The total energy produced by the blast load in the single-degree-of-freedom is:
WImpulse,SDOF =
2
ISDOF
=
2 m SDOF
6.652
= 1724 kip − in.
4.95
2×
386
For this W12 × 53, assuming a uniformly distributed load, the following properties define the structural behavior:
M pl = Zfds =
el =
77.9 × 65
= 422 kip − ft
12
5 422 × 12 × (15 × 12)2
5 M pl L 2
=
= 1.38 in.
48 E I
48
29,000 × 425
410
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The parameters to use in the dynamic calculation are:
8M pl
8 × 422
=
= 225 kips, for the SDOF
L
15
= R yield K L = 225 × 0.5 = 112.5 kips
R yield =
R yield,SDOF
R yield
225
= 163 kips/in., for the SDOF
=
el
1.38
= 163 × 0.5 = 81.5 kips/in.
K =
KSDO F
The period of the system is:
T = 2π
m SDOF
=2×π ×
K SDOF
4.95
= 0.08 sec
386 × 81.5
The column period (0.08 sec) is more than 10 times the load duration (0.0062
sec); hence, the assumption of impulsive load can be assumed. The period of
the column is smaller than half of the beam period (0.08 sec < 0.20
= 0.1 sec);
2
hence, the system is uncoupled and can be modeled separately.
Assuming elastic behavior, the maximum elastic energy that can be adsorbed
by the SDOF system is given by the equation:
W S,el,max =
2
Ryield
,SDOF
2K SDOF
=
112.52
= 77.6 kips − in. 1724 kips − in.
2 × 81.5
This energy is smaller than the energy induced by the impulse; hence, this element will achieve plastic behavior. There is not an economical solution to solve
this problem because of the element’s inability to remain elastic for a directly
applied tributary blast load. But the preliminary assumption that based this design in the rigid behavior of the girt is highly conservative; the maximum load
that this element is carrying comes from the reaction in the girt.
Assuming this reaction is static, the system to solve is defined in Figure 15.12.
The maximum bending moment for this configuration is:
Mx = 2 × 23.6 ×
15
= 236 kips − ft
3
The axial load at the column, based on 30 psf roof dead load, is:
P = 30 ×
50 70
×
= 13.1 kips
2
4
In the following, the capacity properties of this element are calculated to be
used in the interaction axial-bending curves.
EXAMPLES
411
Figure 15.12 Equivalent Static Reaction System
For buckling about the weak axis, the column is assumed unbraced by the
girts. The buckling length is K y L = 1 × 15 = 15 ft; hence, the slenderness is:
Ky L
15 × 12
=
= 72.6
ry
2.48
For the buckling about the strong axis, the column is also unbraced. The buckling
length is K x L = 1 × 15 = 15 ft; hence, the slenderness is:
15 × 12
Kx L
= 34.4
=
rx
5.23
Also,
Cc =
2π 2 E
=
f ds
2 × π 2 × 29000
= 94
65
And the allowable compression stress is calculated by Equation 15-14:
72.62
1−
× 65
2 × 942
Fa =
= 24 ksi
5 3 72.6 1 72.63
+ ×
− ×
3 8
94
8
943
412
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The nominal compression capacity results:
Pu = 1.90AFa = 1.90 × 15.6 × 24 = 711 kips
The Euler stress in the x-axis is:
Fex =
12π 2 E
12 × π 2 × 29000
=
= 129 ksi
23λ2x
23 × 342
In this equation, 12/23 is the factor of safety for buckling. And the unfactored
Euler load in the x-axis results:
Pex =
23
23
AFex =
× 15.6 × 129 = 3857 kips
12
12
The tension capacity is:
Pp = A f ds = 15.6 × 65 = 1014 kips
The plastic moment was calculated before as:
M pl,x = Z x f ds =
77.9 × 65
= 422 kip − ft
12
This value was calculated assuming the member was completely braced, but
this beam is laterally unsupported because the girts cannot carry any axial load
after the blast. Hence, the unbraced flexural strength is:
√
√
L/r y f ds
15 × 12/2.48 × 65
M pl,x = 1.07 −
Mm,x = 1.07 −
3160
3160
×422 = 0.88 × 422 = 371 kip − ft
The first interaction curve that the system should satisfy is:
P
1 × 236
Cmx Mx
13.1
+
+
=
P
13.1
Pu
636
1−
Mmx
1−
× 371
Pex
3857
= 0.015 + 0.64 = 0.65 ≤ 1
where
Cmx = 1
All the other parameters were defined previously.
The second interaction curve to accomplish is:
Mx
P
236
13.1
= 0.56 ≤ 1 for
= 0.013 < 0.15
=
=
M pl,x
422
Pp
1014
EXAMPLES
413
Figure 15.13 Composite Beam Section
This calculation does not consider the reduced stiffness due to the interaction
of bending and compression. For higher compression loads and if the moment
magnifier is bigger than 10 percent, the analysis should consider second-order
effects.
Composite Beam In this example, the composite beam shown in Figure 15.13
will be designed for the roof blast load. The structure is a 25-ft span composite
beam, W14 × 22 with a 51/2 slab with a 3 metal deck. The concrete strength
is 3.0 ksi. For blast design, this strength is multiplied by 1.12. The steel used is
Grade 50, as in the previous examples.
The section properties based on the blast strength for the concrete and the
steel are:
I = 830 in.4
M pl = 395 kip − ft
For rebound purposes, the moment capacity of the steel, based on Z =
33.2 in3 , is:
M pl,steel = Z f dy =
33.2(65)
= 180 kip − ft
12
The load to be used for this example is shown in Figure 15.14.
Finally, assume that the existing dead load is 60 psf.
The total weight for the beam is:
w=
60 × 25 × 6
= 9 kips
1000
The peak load for the 12.2 psi peak blast pressure is:
Fpeak =
12.2 × 12 × 12
× 6 × 25 = 263 kips
1000
414
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
Figure 15.14 Roof Blast Load
Based on the single-degree-of-freedom simplification used for mass and loads
uniformly distributed in the plastic range, the load and the stiffness are multiplied
by the load factor K L = 0.50, and the mass is multiplied by the mass factor
K M = 0.33. Therefore, the parameters used in the discrete system are:
Fpeak,SDOF = 0.5 × 263 = 131.5 kips
wSDOF = 0.33 × 9 = 2.97 kips
The maximum elastic deflection is:
el =
5 395 × 12 × (25 × 12)2
5 M pl L 2
=
= 1.85 in.
48 E I
48
29000 × 830
The parameters to use in the dynamic calculation are:
R yield =
8M pl
8 × 395
=
= 126 kips
L
25
EXAMPLES
415
For the design of this beam, the existing dead load is applied simultaneously
with the blast load and will reduce the beam capacity.
R yield,r educed = 126 − 9 = 117 kips. For the SDOF
Ryield,SDOF = 117 × 0.5 = 58.5 kips
R yield
126
= 68 kips in. For the SDOF
=
el
1.85
= 68 × 0.5 = 34 kips in.
K =
K SDOF
The period of the system is:
T = 2π
m SDOF
= 2π
K SDOF
wSDOF
= 2π
gK SDOF
2.97
= 0.09 sec
386 × 34
For this example, the beam period is less than 10 times the load
duration,0.09 sec < 10(0.0144) = 0.144 sec; hence, the impulse formulation
cannot be used.
For the rebound
R yield,steel =
8M pl,steel
8 × 180
=
= 57.6 kips
L
25
The elastic deflection for the noncomposite steel beam is:
el,steel =
40 M pl,steel L 2
40 180 × 12 × (25 × 12)2
= 3.50 in.
=
384 E Isteel
384
29000 × 199
The maximum negative force is:
RRebound,SDOF = 0.5 × 57.6 kips = 28.8 kips
Therefore, the elastic stiffness at the rebound is:
K Rebound,SDOF =
RRebound,SDOF
28.8
= 8.22 kips in.
=
el,steel
3.50
The period at the rebound is:
T = 2π
m SDOF
K Rebound,SDOF
=2×π ×
2.97
= 0.19 sec
386 × 8.22
416
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
For the rebound, the beam period is greater than 10 times the load
duration,0.19 sec > 0.0144 sec; hence, the impulse formulation can be used for
the rebound response.
The parameters for the load used in the analysis were previously calculated:
wSDOF = 2.97 kips
Fpeak,SDOF = 131.5 kips
The load is applied follow the timing shown in Figure 15.14.
The yield and stiffness were obtained previously as:
Ryield,SDOF = 58.5 kips
K SDOF = 34 kips/in.
And the rebound properties are:
RRebound,SDOF = 28.8 kips
K Rebound,SDOF = 8.22 kips in
Using nonlinear SDOF software, the maximum displacement is (see
Figure 15.15):
SDOF = 1.84 in.
SDOF,Rebound = 3.61 in.
Figure 15.15 SDOF Displacement
EXAMPLES
417
The dead load deflection should be added to the solution obtained by SDOF.
dead =
5 9 × (25 × 12)3
5 W L3
=
= 0.13 in.
384 E I
384 29000 × 830
The maximum deflection results:
max = SDOF + dead = 1.84 + 0.13 = 1.97 in.
Hence, the ductility results:
µ=
1.97
max
=
= 1.06
el
1.85
For the rebound, the ductility is:
µ=
3.61
= 1.03
3.50
For the SDOF solution the element behaves as a composite member and
yields during the rebound. Figure 15.16 shows the force resultant from the SDOF
system.
Figure 15.16 SDOF Force
418
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
The maximum shear force in plastic design can be obtained as the shear reaction that plastifies the section:
V =
R yield
126
=
= 63 kips
2
2
The dynamic shear capacity is:
V p = f dv Aw = 36 × 2.76 = 99.4 kips > 63 kips
where f dv is the blast shear capacity
Aw is the area of the web
Hence, the section can support the shear. The connection should be designed
for this capacity
15.7 DESIGN OF CONNECTIONS
Connections require as much attention as members, especially in blast design.
Failures of steel structures under overload can initiate at the connections of members, as opposed to the members themselves, for many reasons:
r The design of members is often controlled by considerations other than
strength, including deflection, vibration control, and architectural requirements. Such members are often substantially stronger than required to resist the design forces. Connections, however, typically can be designed
based only on considerations of strength. When this is the case, there may
be less overstrength or reserve capacity to match the overstrength of the
members.
r Connections may be controlled by limit states with varying levels of ductility. When less ductile limit states, such as tension rupture or block shear
rupture control, there is less capacity for redistribution of forces and the
mobilization of ductile behavior.
r Connections tend to be of limited size and, therefore, even when exhibiting
ductile behavior, they can only accommodate limited plastic deformation
before reaching their ultimate capacities
It is important in the design of blast-resistant structures, particularly those
expected to be loaded into the inelastic range of behavior, that sudden failure
modes be avoided so that the plastic response of the structure can be mobilized.
It is also important to remember that blast loading may result in load reversal,
REFERENCES
419
Table 15.3 Dynamic Increase Factors for Connection Design
Connection Material
Action
Yield Strength
Ultimate Strength
Plate, A36, A572
Bending/shear
1.3
1.1
Tension/compression/bearing
1.2
1.1
Plate A588
Bending/shear
1.2
1.05
Tension/compression/bearing
1.12
1.05
Bolts
Shear/tension
1.0
1.0
Weld metal
Same as base metals
sometimes in the form of rebound. Connections should be designed with equal
strength under load reversal, unless the following apply:
1. Member strength is limited in one direction of load application by considerations of buckling, as will occur for braces in compression and flexural
members with one flange braced and the other unbraced.
2. Nonlinear dynamic analysis is performed to determine the maximum connection loads in both positive and negative loading applications.
Connections should be designed to develop the full plastic capacity of the supported members, so that the plastic response of the structure can be mobilized in
resisting blast-induced stresses. It should be noted that the dynamic plastic capacity of an element, loaded briefly by impulsive loading, is often greater than the
static plastic capacity. Alternatively, when nonlinear dynamic analysis of structures and structural elements under blast loading is performed, it is permissible
to design the connections for the peak forces obtained from the analysis.
The specified yield and tensile strength of the connection material, including
bolts, welds, angles, and plates, may be increased in accordance with Table 15.3
to account for the Dynamic Increase Factor (DIF). Note that Strength Increase
Factors (SIF), discussed previously for members, are not used in the design of
connections as an additional means of ensuring that connections will be capable
of developing the strength of the member.
REFERENCES
Air Force Engineering and Services Center. 1989. Protective Construction Design Manual. (ESL-TR-87-57). Tyndall Air Force Base, FL: Air Force Engineering and Services
Laboratory.
American Institute of Steel Construction. 2005a. Seismic Provisions for Structural Steel
Buildings. Chicago, IL: American Institute of Steel Construction.
. 2005b. Specification for Structural Steel Buildings. Chicago, IL: American
Institute of Steel Construction.
420
DESIGN CONCEPTS AND MEMBER DETAILING: STEEL
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
ASTM. 2006. Standard Specification for Structural Steel Shapes (A992). West
Conshohocken, PA: ASTM International.
. 2007a. Standard Specification for Pipe, Steel, Black and Hot-Dipped, ZincCoated, Welded and Seamless (A53). West Conshohocken, PA: ASTM International.
. 2007b. Standard Specification for Cold-Formed Welded and Seamless
Carbon Steel Structural Tubing in Rounds and Shapes (A500). West Conshohocken,
PA: ASTM International.
. 2008. Standard Specification for Carbon Structural Steel (A36). West Conshohocken, PA: ASTM International.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5−1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2000. Prestandard and Commentary for the
Seismic Rehabilitation of Buildings (FEMA 356). Washington, DC: Federal Emergency Management Agency.
Gilsanz, Ramon et al. 2009. Defensive Design of Structural Steel Buildings (AISC Design
Guide No. 27), Chapters 6 and 7. Chicago, IL: American Institute of Steel Construction.
Shipe, James A. and Charles J. Carter. 2003. Defensive design. Modern Steel Construction 43 (11): 25–31.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures to Conventional Weapons Effects (FOUO) (UFC 3-340-01).
Washington, DC: U.S. Department of Defense, Unified Facilities Criteria
Program.
. 2008. Structures to Resist the Effects of Accidental Explosions (UFC 3-34002). Washington, DC: U.S. Department of Defense, Unified Facilities Criteria Program.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
16
Blast-Resistant Design
Concepts and Member
Detailing: Masonry
Shalva Marjanishvili
Prior to the twentieth century, unreinforced masonry was the primary building
material for both residential and commercial construction. As a result, unreinforced masonry structures constitute a large portion of existing buildings, many
of which have historical and architectural importance.
Beginning in the early 1980s with terrorist attacks in Lebanon and Kuwait,
high-profile and high-risk civilian federal buildings have been designed to resist
the effects of explosive attacks. With each major attack on U.S. interests, more
attention has been paid to this low probability/high consequence threat scenario.
As a result, recent work has been dedicated to the design and analysis of masonry
structures to resist air-blast loads.
When a wall is first subjected to an air-blast load, it experiences out-of-plane
flexure, producing tensile strains on the interior face of the wall and compressive
strains on the exterior. Once the maximum positive deflection associated with
the out-of-plane flexure has been attained, the wall will begin to vibrate due to
rebound forces created by negative blast pressures. The wall undergoes negative
deflection and curvature, causing tensile strains to develop on the exterior face
of the wall, and increasing shear stresses at wall supports.
Concrete masonry unit construction is primarily used for walls. These walls,
when properly designed, will provide economical resistance to relatively low
air-blast pressures. Limitations of masonry walls include development of large
deformations and rebound response.
Most masonry design codes advocate elastic design to specified stress levels, even for dynamic loads. This approach is uneconomical and potentially unsafe because of its inability to predict structural response beyond the allowable
stress limit. Masonry walls, when subjected to air-blast loads, are expected to undergo large post-elastic response, and therefore, they should be designed using
the strength design methods.
Below are two primary assumptions made when assessing flexural strength of
reinforced masonry:
421
422
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Assumption 1.
Singly reinforced masonry units rely on composite action between compression in concrete and tension in reinforcement.
a. Plain sections remain plain during bending. At large deflections, significant deviation from this assumption may occur. This can be corrected by
including both bending and shear deformations in the analysis.
b. There is a perfect bond between grout and reinforcement.
c. Concrete tension strength is neglected.
d. Concrete stress block is idealized as a rectangle.
e. Reinforcement stress/strain is idealized as elastic-perfectly plastic with appropriate strain-rate hardening effects.
f. Flexural strength is attained when either tension reinforcement or extreme
fibers of concrete reach their ultimate strains.
Assumption 2.
Under extreme loading such as air blast, it is expected that the compressive
strength of masonry will be exceeded. When this happens, the resistance is provided purely by compression and tension in the reinforcement. This approach requires that masonry be reinforced with two curtains of reinforcement. Therefore,
doubly reinforced masonry units, under this assumption, rely on compressive and
tension forces in the reinforcement only.
a. Plain sections remain plain during bending.
b. Reinforcement is assumed to be held in its original position, as if the masonry were undamaged.
c. Concrete tension and compression strengths are neglected.
d. Reinforcement stress/strain relationship is idealized as elastic-perfectly
plastic with appropriate strain hardening effects.
e. Maximum flexural strength is attained when either curtain of reinforcement yields either in tension or in compression.
Assumption 1 is appropriate only for level of protection (LOP) I, the superficial damage level. For other damage levels (LOP II–LOP IV), it is necessary to
use Assumption 2. It is worth noting that reinforced masonry will often behave
under Assumption 2.
When a masonry wall is subjected to relatively small air-blast pressures with
low performance requirements (LOP I), ACI 530-08, Section 3.3.5 (American
Concrete Institute 2008) can be used to estimate out-of-plane moment capacity
of the wall, with the following modifications:
1. φ strength reduction factors shall be taken as one (i.e., φ = 1.0).
2. Yield strength of reinforcement shall be taken as dynamic (i.e., f y = f dy ).
3. Compressive strength of masonry shall be taken as dynamic (i.e., f m =
).
f dm
GENERAL CONSIDERATIONS
423
UFC 3-340-02 (Unified Facilities Criteria 2008) suggests the following equation for estimating ultimate moment capacity under Assumption 2:
Mu = As f dy dc
(UFC 3-340-02, Equation 6.2)
where: As = area of joint reinforcement at one face
f dy = dynamic yield strength of reinforcement
dc = distance between centroids of compression and tension reinforcement
16.1 GENERAL CONSIDERATIONS
Reinforced masonry subjected to service design loads (excluding air-blast) can
be designed using both allowable and strength design methodologies. For airblast design, only strength design methodology should be used, since reinforced
masonry elements subjected to air-blast loads are likely to experience large nonlinear deformations with relatively large strain and ductility demands. Therefore,
it is necessary to employ a design methodology that can account for response
well beyond linear limit states. Working stress design, although much simpler
then strength design, only accounts for response below nonlinear range, and implies that all reinforced masonry should be designed linearly. Furthermore, airblast design requires that structural elements respond in a ductile manner, with
large nonlinear deformations. For these reasons, allowable stress design should
not be used for air-blast design of reinforced masonry.
Although unreinforced masonry should not be used for new construction,
there are many historic buildings with existing unreinforced masonry walls.
When existing unreinforced masonry walls cannot be replaced, the stability of
existing unreinforced masonry walls should be assessed. The stability assessment should assume that the wall is simply supported if flexible supports are
provided. It should be permissible to assume arching action in the masonry wall
if rigid supports are provided.
There will be cases when air-blast design requirements contradict or largely
exceed the conventional masonry design required by building codes. The design
engineer should verify that the final masonry design meets all design load requirements, including blast design requirements. If the blast design requirements
countermand or contradict other masonry design constraints, the owner and architect should be appraised of the issue by the design professional, and a decision
should be reached by the group in order to resolve the conflicting requirements.
The design engineer should direct the detailing on the design documents such
that all requirements of the blast-resistant design reworking and renovation of an
existing structure are met. The design engineer should verify that the design documents clearly detail any required repairs of the existing structure, modifications
to the existing structure, or further work to the existing structure that is necessary
to achieve the LOP required of the modified existing structure.
424
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
16.1.1 Masonry
Maximum masonry strength, f m, should not exceed 6,000 psi for any LOP in
new construction. This strength limitation is intended to preclude nonductile failure modes, such as shear and/or crushing.
Minimum masonry strength, f m, should not be less than 1,500 psi for any
LOP in new construction.
The design engineer should determine the material properties of existing masonry structural elements to ensure proper design to achieve the LOP required
for the existing masonry structure.
UFC 3-240-02 suggests the following equations to determine effective moment of inertia:
Ieff = Ia =
In + Icr
2
Icr = 0.005bdc3
(UFC 3-340-02, Equation 6-6)
(UFC 3-340-02, Equation 6-7)
While ACI 530-08 suggests that Ieff is approximately equal to one-half of
gross moment of inertia, it also suggests the following equation for a more accurate estimate of effective moment of inertia:
Mcr 3
Mcr 3
≤ In ≤ 0.5Ig
+ Icr 1 −
Ieff = In
Ma
Ma
(ACI 530-08, Section 3.5.3)
This handbook recommends the ACI approach when using Assumption 1, and
the UFC approach when using Assumption 2. This recommendation is based on
the understanding that the UFC document covers reinforced masonry response
due to extreme loading conditions, while ACI-350-08 is concerned mainly with
elastic response.
16.1.2 Reinforcement
All reinforcement should comply with ASTM A615 (Grade 60) or A706 for all
LOPs in new construction (ASTM 2006, ASTM 2008a).
It is desirable to use reinforcement with high ductility demand capability. Currently only ASTM A706 reinforcement is considered to have high ductility capacity. However, according to UFC 3-340-02, it is sufficient to use ASTM A615
(Grade 60).
The modulus of elasticity of reinforcement should be taken as required by
ACI 350-08.
At strain rates characteristic of air-blast response, reinforcing steel exhibits
a significant increase in yield strength above static test values. For 60-ksi
reinforcement, UFC 3-340-02 recommends a dynamic increase factor of 1.17,
GENERAL CONSIDERATIONS
425
that is:
f dy = 1.17 f y = 1.17 × 60 ksi = 70.2 ksi
Reinforcement splices should be limited to zones that remain elastic during
blast response. Furthermore, not more than 50% of splices should be located
along a single line. That is, splices should be staggered along the length or height
of the wall and away from potential yield lines.
Splices should develop full tensile capacity of reinforcement and should conform to ACI 350-08. It is possible to use mechanical splices; however, the use of
mechanical splices is limited by the geometric constraints of concrete masonry
units. Use of splices should be limited.
16.1.3 Mortar
Mortar is a mixture of cementitious materials, aggregates and water. All mortars
should comply with ASTM Standard C270 (ASTM 2000). Mortars with lime
should be avoided, as lime decreases the strength of mortar.
16.1.4 Grout
Grout provides bonding of longitudinal reinforcement and adds strength, rigidity, and mass to the wall system. All grouts are required to conform to ASTM
Standard C476 (ASTM 2008b). It is recommended that blast masonry walls be
fully grouted.
16.1.5 Construction Methods
Two commonly used types of masonry construction are running bond and stack
bond.
With running bond, masonry units are normally staggered by at least onequarter of the length of the unit. With typical 8-in. masonry units, the joints are
offset by at least 4 in., and since the length of the units is 16 in., it is common to
offset units by half the length of the unit.
With stack bond, joints in each successive course line up vertically. This type
of construction is not desirable in air-blast-resistant construction, since there is
little or no continuity across the head joints (the head joint is the joint that runs
up and down the wall). The only continuity across the head joints is provided by
longitudinal reinforcement and grout.
Conventional construction methods of reinforced masonry walls include running bond and stacked bond with the maximum spacing of vertical reinforcement
at 48 in. on center (o.c.). This implies that not every cell is grouted.
For air-blast-resistant design, it is recommended that every cell be grouted
and at least one reinforcing bar be placed in each concrete masonry unit. For
example, with 16-in. masonry units the maximum reinforcement spacing should
426
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
not exceed 16 in. o.c., and every cell should be grouted. Since concrete masonry
units (CMUs) are of modular construction with a modular dimension of 8 in.,
reinforcement must be placed in increments of 8 in.
When designing with reinforced masonry, dimensional constraints need to be
accounted for as follows:
1. Actual dimensions of masonry units are 3/8 in. less than the nominal dimensions.
2. Reinforcement should be placed at increments of 8 in. o.c. (i.e., 8 in.,
16 in.).
3. Since the largest CMU block has a length of 16 in., it is recommended that
at least one reinforcing bar be placed in each masonry unit, resulting in a
maximum spacing of 16 in. o.c.
4. Masonry unit wall thickness will dictate the minimum cover dimension.
It may be very difficult to achieve double curtain reinforcement in 4-in.
masonry units.
5. Double curtain reinforcement can be achieved with 8-in. masonry units by
staggering longitudinal reinforcement as shown in Figure 16.9
16.2 FAILURE MODES
Failure modes of masonry walls are categorized as either ductile or brittle.
Ductile failure is preferred, and is generally characterized by bending failure
caused by yielding of tensile reinforcement. Brittle failure modes include shear,
buckling, and crushing. Figures 16.1 though 16.3 are photos depicting failures of
masonry construction. Figure 16.1 shows infill masonry wall performance when
structural collapse is initiated. Figure 16.2 depicts infill masonry wall failure
without collapse of bearing elements. In this case, the masonry wall could
potentially become heavy debris thrown into the occupied rooms. Both Figures
16.1 and 16.2 show masonry wall performance under extremely high air-blast
loads, while Figure 16.3 shows reinforced masonry wall performance under
moderate to low air-blast pressures. Although the masonry wall shown in Figure
16.3 experienced some cracking and stucco spalling, it performed satisfactorily.
These figures illustrate how reinforced masonry walls could provide economical
protection under relatively moderate air-blast pressures.
Table 16.1 correlates qualitative damage description with the damage levels
superficial, moderate, heavy, and hazardous and their corresponding levels of
protection from IV though I.
The first step in air-blast design of masonry walls is to establish the desirable performance level in terms of acceptable damage levels, as shown in Table
16.1. Then, dynamic analysis is conducted to calculate damage indicators, as described under each column for corresponding damage levels. If it is determined
that the damage indicators are exceeded for a desired performance level, the
FAILURE MODES
Figure 16.1 Failure of Building with Masonry Infill Walls
Figure 16.2 Failure of Masonry Infill Wall
427
428
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.3 Crack Pattern on Masonry Infill Wall
wall must be redesigned and re-analyzed. Thus, air-blast design is an iterative
process.
To simplify air-blast analysis and design procedures, often a single-degree-offreedom (SDOF) analysis is conducted with a predetermined deflected shape of
the wall. The results of this SDOF analysis are usually obtained as ductilities and
support rotations. These calculated ductilities and rotations are then compared to
allowable values, to determine damage limits (see Chapter 3, section 3.5 of this
book for allowable ductilities and rotations for reinforced masonry).
Table 16.1 also lists performance indicators for unreinforced masonry. Although unreinforced masonry is not recommended in new construction, there are
many existing buildings with unreinforced masonry facades which may require
evaluation under air-blast loads.
16.2.1 Flexure
Flexure provides a ductile failure mode. This is the preferred failure mechanism
and occurs in properly designed and supported masonry. This can be achieved by
designing masonry for the balanced condition whereby reinforcement will yield
just before the masonry concrete begins crushing. Figure 16.4 depicts typical
deflected shape due to flexural response.
When air-blast pressures are relatively low, it may be possible to design a reinforced masonry wall with just one curtain of reinforcement located in the middle.
429
Limit State
Multiple
Multiple
Element
Reinforced
Unreinforced
Not permitted
Masonry is effective in
resisting moment and
shear. The cover over the
reinforcement on both
surfaces of the wall
remains intact. System
should essentially remain
elastic.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.003.
Superficial
LOP IV
Significant spalling and
scabbing of masonry.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.003.
Significant spalling and
scabbing of masonry.
No connection or shear
failure.
Compressive strain in
masonry does not exceed
0.007.
Masonry is completely
disengaged.
Significant spalling and
scabbing of masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry exceeds 0.010.
The concrete cover over the
reinforcement on both
surfaces of the element is
completely disengaged.
Equal tension and
compression
reinforcement that is
properly tied together are
required to resist moment
Limited spalling and
scabbing of masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry should not
exceed 0.010.
Masonry is crushed and not
effective in resisting
moment; however, shear
resistance remains though
interlocking. Compression
reinforcement equal to the
tension reinforcement is
required to resist moment.
The cover over the
reinforcement on both
surfaces of the wall
remains intact.
No spalling or scabbing of
masonry.
Axial load-bearing elements
maintain bearing capacity.
No connection or shear
failure.
Compressive strain in
masonry should not
exceed 0.007.
Not permitted
Hazardous
LOP I
Heavy
LOP II
Moderate
LOP III
Table 16.1 Qualitative Damage Expectations for Masonry
430
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.4 Flexural Response Mode of Reinforced Masonry
In this case, Assumption 1 is used to determine the bending capacity of the masonry. The wall should still be designed so that tensile yielding of reinforcement
occurs before the compressive block of masonry starts crushing. This will result
in a relatively weak but ductile wall.
When air-blast pressures are relatively high (in excess of 10 psi), it may be
necessary to design the reinforced masonry wall with a double curtain of reinforcement, using Assumption 2. In this case, it is desirable to place equal
amounts of reinforcement on each side of the wall. Furthermore, it may be necessary to increase the thickness of the masonry wall to accommodate two curtains
of reinforcement. This can, however, result in a stronger and more rigid wall,
which may become susceptible to diagonal and direct shear failure. Thus, there
FAILURE MODES
431
should be an analysis check, and if necessary additional design iterations, to
confirm that the wall is designed in such a way that it reaches the flexural capacity limit before the shear capacity limit, when subjected to load distribution
similar to air-blast load.
16.2.2 Diagonal Tension Shear
Diagonal tension shear failure occurs when a masonry wall response reaches
the diagonal tension resistance limit before the bending capacity is exhausted.
Figure 16.5 depicts typical deflected shape corresponding to diagonal tension
shear response. Diagonal tension failure exhibits very little ductility capacity
and therefore is brittle in nature. Because of this, diagonal tension failure should
Figure 16.5 Diagonal Shear Response Mode of Reinforced Masonry
432
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
be avoided in blast-resistant masonry walls. One method to avoid diagonal tension failure mode is to decrease the flexural capacity of the wall in such a way
that the wall undergoes flexural yielding before shear capacity is exhausted.
However, the flexural resistance of the wall should not be reduced when doing so will compromise the desired level of protection. Alternatively, the diagonal tension capacity of the masonry wall can be increased by increasing the
thickness of the masonry, by grouting more cells, or by placing shear reinforcement. Placement of shear reinforcement is feasible only with two curtains of
reinforcement. If shear reinforcement is required, it should be placed in such a
way that it crosses expected diagonal shear crack between the two curtains of
reinforcement.
Diagonal tension capacity for out-of-plane bending of reinforced masonry
walls can be assessed using ACI-350-08 strength design procedures.
16.2.3 Direct Shear
Direct shear occurs when explosion occurs very close to the wall and when mortar joints are too weak to provide shear stress flow continuity across the masonry
units. Figure 16.6 depicts typical deflected shape consistent with diagonal shear
response. Direct shear may be of concern when scaled distance is less then 4. At
this standoff it is expected that air-blast pressures will be of very high magnitude
and extremely short duration. If the air-blast pressures are of extremely short duration, there will be no time for the masonry wall to respond in flexure mode,
and direct shear response will dominate.
Direct shear capacity depends on friction resistance across the mortar joint
and dowel action of the longitudinal reinforcement. Direction shear strength generally exceeds diagonal shear strength. The direct shear failure mode is almost
always very brittle and should be avoided. One way to avoid direct shear failure
is to increase the standoff between the masonry wall and the expected explosion
source. Furthermore, since reinforced masonry walls do not perform well under
extremely high air-blast pressures, direct shear failure modes should be avoided
whenever possible.
16.2.4 Breach and Spall Phenomena
Breach occurs when a weapon is placed directly against or very close to the
masonry wall surface. Detonation at this close proximity causes shattering of
the masonry units. Breach analysis is often conducted using computational fluid
dynamic codes with appropriate equations of state, or by experimental studies.
Breach effects can be mitigated by masonry wall thickness, proper confinement,
and application of anti-spall laminates such as fiber-reinforced polymer (FRP).
Spall typically occurs on the back side of the wall when a weapon is
placed at the breaching distance but does not cause full breach. Spall occurs
due to a compression wave traveling through the thickness of the masonry.
FAILURE MODES
433
Figure 16.6 Direct Shear Response Mode of Reinforced Masonry
When reflected from the back surface of the wall, this causes through-thickness
tensile stresses at the back face that can exceed the tensile strength of the
masonry.
Both breach and spall will cause fragmentation on the back side of the wall.
Spall can be mitigated by application of shielding (laminate material) such as
FRP or so-called “catchment systems” such as geotextile fabric to stop fragments. Laminate shielding is permanently adhered to the wall surface (on the
protected side of the wall) and may act compositely with the masonry, while
catching systems are basically a curtain or fabric located on the protected side
of the wall to catch debris generated by failed masonry. Both systems provide protection by the development of membrane resistance, and therefore they
should be connected to supports (normally floor slabs) in a way that allows full
434
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
development of the membrane capacity of the base material. This is done to avoid
premature failure of the system. Fragmentation effects are discussed in the detail
in Chapter 8 of this book.
16.3 REINFORCED MASONRY DETAILING
The detailing guidelines presented in this section are intended to promote the
desired structural behaviors discussed in Chapter 9.
Since masonry walls resist air-blast loads primarily though nonlinear deformations, it is important that wall supports develop the full flexural capacity of
the wall. Reinforced and unreinforced walls can be analyzed using single-degreeof-freedom (SDOF) methods or pressure-impulse (P-I) diagrams. Axially loaded
walls should be designed considering slenderness effects.
Non-load-bearing partition walls should be reinforced and grouted, as required, to resist the applied blast loads. Alternatively, debris catch systems
should be used to minimize the debris impact hazard. The effectiveness of the
debris catch systems should be determined through applicable explosive test data
or advanced finite element analysis.
The design shear forces should not be less than the shear forces associated
with the ultimate flexural strength of the element. The shear capacity of reinforced masonry wall structures can be increased by placement of shear reinforcement. Shear reinforcement should be detailed in accordance with UFC 3-340-02.
Masonry wall structures should be sufficiently reinforced to resist the in-plane
and out-of-plane response to dynamic blast loads. It should be permissible to
consider arching action for the evaluation of wall response and to use empirical P-I charts based on masonry wall tests; however, the impulse asymptote
“layover” (the “bump” at the extreme asymptote) associated with the beneficial
effects of negative phase loading may be nonconservative, and should not be considered for design. The use of P-I charts should be limited to flexural modes in
response to external blast loads and to wall types that are within the P-I database.
Diagonal tension and direct shear capacity must be verified independently. Walls
should be designed to develop the shear forces associated with plastic hinging.
Connections should be designed, detailed, and constructed such that the governing failure mode of the connection is a ductile mode related to yielding of the
steel elements of the connection, such as angles, plates, and rebar. Fracture of
bolts and welds should not be the governing failure mode of connections.
This section applies to non-load-bearing masonry walls only. Reinforced concrete masonry units should allow full development of the reinforcing steel. Reinforced concrete masonry unit walls should be designed as either one- or two-way
systems using cell and joint reinforcement. Shear resistance can be provided by
the concrete masonry unit in combination with steel reinforcement, although the
inclusion of this behavior should be considered in light of the assumptions identified at the beginning of this chapter.
REINFORCED MASONRY DETAILING
435
16.3.1 General
All concrete masonry units should be fully grouted for LOP III and LOP IV.
All lap splices should be tension lap splices, as specified in ACI 530 and ACI
530.1 (American Concrete Institute 2008).
Use of mechanical splices should be limited to zones that remain elastic during blast response, unless the mechanical splices can be shown to develop the
expected tensile strength of the bar under conditions generated by blast loads.
All mechanical splices should meet or exceed the specifications in ACI 530 and
ACI 530.1.
Use of welded splices should be limited to zones that remain elastic during
blast response, unless the welded splices can be shown to develop the expected
tensile strength of the bar under conditions generated by blast loads. All welded
splices should meet or exceed the specifications in ACI 530 and ACI 530.1.
16.3.2 Longitudinal Reinforcement
For air-blast-resistant design, it is desirable to place longitudinal reinforcement in each cell in fully grouted concrete masonry units. This will result in
reinforcement placed at 8 in. o.c. This detailing approach is beneficial since every grouted cell adds weight, rigidity, and shear strength, and closely spaced bars
may reduce the size of fragments should the wall fail under air-blast loads. Figures 16.7 though 16.10 provide sample detailing options for reinforced masonry.
16.3.3 Horizontal Reinforcement
A bond beam course with two horizontal reinforcing bars should be placed all
around the cap course of the wall. Bars should be spliced and the course grouted,
per requirements of ACI 530.
All new construction should have a bond beam course with a minimum of two
horizontal reinforcing bars. Reinforcing bars should be located at the level of the
lowest attachment of the masonry wall to the floor diaphragm. Bond beam should
be placed at any horizontal diaphragm or floor system supported by the wall. If
the attachment band in the diaphragm is thicker than the bond beam course, then
the bond beam course should be placed on the bottom of the diaphragm, and
all cells in courses to the top of the attachment band should be fully grouted.
Bars should be spliced and the course grouted per requirements of ACI 530. See
Figure 16.11 for examples of bond beam reinforcement.
A bond beam course with two horizontal reinforcing bars should be placed
at all lintel locations; the bar should be either hooked around vertical reinforcement on either side of the opening or continuous along the wall, with the course
grouted per requirements of ACI 530.
Control joints for all LOPs should be designed, at a minimum, with one
vertical bar in a fully grouted cell column on each side of the control joint.
436
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Figure 16.7 Fully Grouted Reinforced Masonry Wall
Figure 16.8 Singly Reinforced Masonry Wall
Figure 16.9 Masonry Wall Doubly Reinforced by Staggering Longitudinal Reinforcement
REINFORCED MASONRY DETAILING
Figure 16.10 Doubly Reinforced Masonry Wall – Two Bars per Each Cell
(a)
(b)
Figure 16.11 Examples of Bond Beam
437
438
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
Code-approved control joint block and expansion material systems should be
used as necessary in all blast-resistant masonry systems.
The design professional should determine the appropriate locations for control
joints, to the extent that their locations impact blast resistance.
The design professional should specify continuity and/or end conditions of
horizontal bars and bond beams at control joints.
16.3.4 Walls
All elements constructed to LOPs I and II should have at least one continuous
vertical reinforcing bar in a fully grouted cell column, with spacing not to exceed
48 in. o.c.
All elements constructed to LOP III and IV should have at least one continuous vertical reinforcing bar per block in a fully grouted cell column, with spacing
not to exceed 16 in. oc.
In addition, at least one vertical bar should be placed in a fully grouted cell
column at all inside or outside corners, alongside any openings, and in the cell
column on each side of control joints, if they are used in the structure
All elements constructed to LOP III and IV should have a minimum vertical
reinforcement ratio of 0.0025.
Bearing Walls Bearing walls are subjected to constant axial load in addition to
blast loading. The presence of axial load in reinforced masonry increases bending
capacity up to a so-called balanced point on a P/M diagram, after which axial
load causes rapid deterioration of bearing wall air-blast-resistance capacity, due
to buckling. Additionally, the presence of axial load induces larger stresses on
block masonry under compression, and may cause the masonry wall to fail in
concrete crushing mode rather than through the yielding of the reinforcement
(i.e., compression-controlled section).
Careful attention must be paid during the structural analysis of bearing walls
to avoid these undesirable failure modes (crushing and buckling).
Slender Walls Slender walls normally carry some axial load, which affects the
overall resistance of the wall and should be accounted for in the analysis. All
slender wall design should follow ACI 350-08, Section 3.3.5
Infill Walls Infill masonry walls typically act as partitions and are supported by
floor slabs or beams. If properly designed they can provide physical separation
and protection against air-blast waves and fragments.
16.3.5 Support Connections
Since masonry walls are primary blast-load-resisting systems, and since masonry
walls are normally expected to respond to air-blast load in nonlinear regime,
UNREINFORCED MASONRY
439
it is necessary to provide sufficient support to allow masonry walls to develop
full plastic mechanism. This can be achieved by fully developing longitudinal
reinforcement into the supports.
16.4 UNREINFORCED MASONRY
Unreinforced masonry, due to its brittle nature, is not recommended in new construction. However, there is a vast amount of unreinforced masonry in existing
and historic structures. Sometimes, existing reinforced masonry walls contain
reinforcement that provides negligible improvement over the unreinforced masonry. Lightly reinforced masonry walls of this type are usually analyzed using
the principles of unreinforced masonry discussed in this section.
Methods have been established for assessing the stability of unreinforced masonry walls under blast loading. The phenomenon of arching action is one of the
primary mechanisms by which unreinforced masonry resists complete failure.
However, even slight in-plane vertical movements of “rigid” supports negate the
large benefits predicted by the assumed arching action.
16.4.1 Performance Evaluation
When an external explosion imposes an air-blast load on a heavy external wall
system, the air-blast pressure must overcome the inertial forces of the wall itself before putting the system into motion and causing damage. In the case of
thick, unreinforced masonry walls, these forces are very large, and therefore the
resulting damage is less than one might intuitively imagine.
Informative data for masonry can be obtained from the Concrete Masonry
Unit Database Software (CMUDS) program. The full-scale equivalent dimensions for the data have spans of 8 to 11 feet and thicknesses of 6 to 8 inches.
CMUDS also provides data on reinforced masonry, with reinforcement ratios
ranging from 0.15% to 00%, although most of the data have reinforcing ratios from 0.15% to 0.30%. The data include full-scale shock tube testing and
high explosives testing on full- and quarter-scale walls. Elastic-perfectly plastic analysis should not be used for unreinforced masonry elements, as this will
lead to a nonconservative calculation of the blast load that such elements can
resist.
Wall Analysis A common configuration for windows in historic unreinforced
masonry buildings, especially those constructed in the late 1800s and the early
1900s, is to have two or three windows per structural bay. This creates masonry
piers between individual window openings. These piers tend to be critical wallsupport elements.
It is best to begin calculations assuming that the windows can be anchored
to the walls at all four sides, distributing the forces from the windows to the
440
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
supporting walls at the head, sill, and jambs. Walls should have sufficient dynamic capacity to allow window glazing to break first. This approach, commonly
known as balanced design, has been shown to reduce the weight (the size and the
amount) of debris that may enter the occupied space. There is a direct correlation
between injuries and size and velocity of debris.
Two computational models for analyzing wall response are described below.
Simply-Supported Wall A conservative approach to modeling the resistance of
an unreinforced masonry wall for an air-blast load is to represent the wall as a
slender member subjected to a uniformly applied dynamic load. It is assumed
that there is no contribution of axial compressive forces at the top and bottom
of the wall, and the wall is analyzed as a simply-supported beam. The flexural resistance of the wall is dependent on the compressive stresses from its self
weight, and the tensile strength as defined by the mortar’s modulus of rupture.
Because unreinforced masonry has little tensile resistance, the flexural capacity
of the wall is minimal, resulting in extensive cracking and brittle failure at the
interior face of the wall.
Rigid Supports Another model, based on UFC 3-340-02 (the recognized
air-blast-resistance design manual throughout the United States), assumes
that the top and bottom supports are completely rigid and provide restraint
against elongation. Strength comes from compressive blocks that form at the
top, bottom, and mid-height hinge locations of the wall, behavior known as a
compressive membrane phenomenon. This model also assumes that the mortar
joints will be cracked, disregarding what little flexural tensile strength remains
in the masonry. The deflected wall “wedges” itself between floor slabs and also
forms a plastic hinge at mid-height. This model takes into account that masonry
infill walls will often have a gap at the top, either from a mortar joint or from a
design that accounts for flexure of the floor systems above.
16.4.2 Retrofit Recommendations
Retrofit masonry walls can be costly, but may be unavoidable due to various
constraints such as the historic value of a building. Masonry wall retrofits are
usually required when an existing wall cannot be replaced and does not meet the
desired performance level of protection. Retrofit recommendations provided in
this section apply equally to unreinforced and reinforced masonry walls.
Catchment Systems Or Curtains Depending on the performance intent, an alternative to structural retrofit of unreinforced masonry walls is to design geotextile fabric blast curtain located on the inside face of a non-blast-mitigating masonry wall. This curtain will provide protection from masonry fragments. The
geotextile fabric and connections should be designed for membrane action of the
fabric. The fabric connections should transfer loads into the floor system without
UNREINFORCED MASONRY
441
causing fabric tears from stress concentrations. The performance of a specified
geotextile fabric and its connections should be verified by blast testing. Generally, blast testing, in addition to demonstrating acceptable performance for design
loads, should also demonstrate the post-failure behavior of the retrofit system, to
ensure that the failure of the system does not create brittle and hazardous failure.
Secondary Walls A secondary blast-mitigating wall system installed on the
inside face of an existing wall system, in addition to being designed to withstand
fragmentation and impact effects (due to the frangible wall in front), should also
be designed as a new wall, in accordance with this chapter.
Steel Supporting Frames The most straightforward retrofit is to install steel
frames, interior to the existing walls, that support the wall and the window system. This solution is appropriate where the wall piers are not able to support the
additional air-blast load imposed by the windows. This retrofit is often unacceptable to architects, as it either creates an uneven wall surface or requires reducing
the overall room size by furring out the walls.
Shotcrete Walls In some cases, other loading condition requirements may be
combined with air-blast requirements to develop multipurpose solutions. Where
progressive collapse or seismic design requirements would benefit from shotcrete
walls, this may be used as an air-blast retrofit of unreinforced masonry walls as
well. Window systems could be mounted in the new concrete wall system. This
would be an especially useful technique for relatively larger blast loads.
Fiber-Reinforced Polymer (FRP) Another retrofit method is to apply vertical FRP composite strips along the inside face of the masonry wall. FRP
composites have significant tensile strength in the direction of the fibers and
can be highly effective in increasing the wall’s out-of-plane flexural capacity. Additional advantages of this method are that it is relatively unobtrusive
to the architectural detailing of the wall and existing structure, and its application process is not labor-intensive, requiring little disruption to building
occupants.
The flexural capacity of an FRP-retrofitted unreinforced masonry wall can
be assessed using basic moment-equilibrium relations, as with a steel-reinforced
masonry wall. The tensile strength of the masonry is disregarded, and all tensile
resistance is assumed to be provided by the FRP. The flexural resistance can
then be computed as a function of the compressive strength of the masonry, the
tensile strength of the FRP, the thickness of the masonry panel, and the value of
the applied axial force.
Blast tests have demonstrated that it is unlikely that the wall’s predominant
failure mode will be failure of the FRP. Also, tests of unreinforced masonry
units to out-of-plane air-blast loading have demonstrated that shear failure at the
supports is a predominant failure mechanism. The FRP reinforcing provides no
442
DESIGN CONCEPTS AND MEMBER DETAILING: MASONRY
additional shear resistance to the wall section, and if connections at the wall supports do not allow transfer of excessive shear forces to other structural elements,
shear failure at the wall supports can occur. Also, connections at wall supports
are instrumental in preventing out-of-plane flexural collapse towards the front
side of the wall due to rebound forces.
REFERENCES
American Concrete Institute. 2008. Building Code Requirements for Masonry Structures
and Commentary (ACI 530-08/ACI 530.1-08/ASCE 5-08/TMS 602-08). Farmington
Hills, MI: American Concrete Institute.
American Society of Civil Engineers, Petrochemical Committee, Task Committee on
Blast Resistant Design. 1997. Design of Blast Resistant Buildings in Petrochemical
Facilities. New York: American Society of Civil Engineers.
ASTM. 2000. Standard Specification for Mortar for Unit Masonry (C270). West Conshohocken, PA: ASTM International.
. 2006. Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement (A706). West Conshohocken, PA: ASTM International.
. 2008a. Specification for Deformed and Plain Billet-Steel Bars for Concrete Reinforcement (A615). West Conshohocken, PA: ASTM International.
. 2008b. Standard Specification for Grout for Masonry (C476). West Conshohocken, PA: ASTM International.
Brandow G. E., G. C. Hart, and A. Virdee. 1997. Design of Reinforced Masonry Structures. Citrus Heights, CA: Concrete Masonry Association of California and Nevada.
Departments of the Army, the Navy and the Air Force (DOANAF). 1990. Structures to
Resist the Effects of Accidental Explosions (Department of the Army Technical Manual
TM 5−1300). Washington, DC: Departments of the Army, the Navy and the Air Force.
Federal Emergency Management Agency. 2003. Primer for Design of Commercial Buildings to Mitigate Terrorist Attacks (FEMA 427). Washington, DC: Federal Emergency
Management Agency, Department of Homeland Security.
Paulay, T. and M. J. N. Priestley. 1992. Seismic Design of Reinforced Concrete and Masonry Buildings. Hoboken, NJ: John Wiley and Sons.
Protective Design Center. 2008. Single Degree of Freedom Structural Response Limits for
Antiterrorism Design (PDC TR-06-08). Omaha, NE: U.S. Army Corps of Engineers,
Protective Design Center.
Unified Facilities Criteria Program. 2002. Design and Analysis of Hardened Structures
to Conventional Weapons Effects (UFC 3-340-0). Washington, DC: U.S. Department
of Defense, Unified Facilities Criteria Program. For Official Use Only.
Unified Facilities Criteria Program. 2006. DoD Minimum Antiterrorism Standards for
Buildings (UFC 4-010-01). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program.
Unified Facilities Criteria Program. 2008. Structures to Resist the Effects of Accidental
Explosions (UFC 3-340-02). Washington, DC: U.S. Department of Defense, Unified
Facilities Criteria Program. Supersedes TM 5-1300.
UNREINFORCED MASONRY
443
U.S. Army Corps of Engineers. 1999. Airblast Protection Retrofit for Unreinforced
Concrete Masonry Walls (ETL 1110-3-494). Washington, DC: U.S. Army Corps of
Engineers.
U.S. Department of State, Overseas Buildings Operations. 2002. A&E Design Guidelines
for U.S. Diplomatic Mission Buildings. Washington, DC: U.S. Department of State,
Overseas Building Operations.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
17
Retrofit of Structural
Components and Systems
John E. Crawford and L. Javier Malvar
17.1 INTRODUCTION
Structural components such as columns, beams, floors, and walls are likely to be
vulnerable to blast loads. This is important because the resulting damage to these
components could produce partial or complete collapse of the structural system,
or produce large volumes of injurious debris. For example:
r The blast at the Alfred P. Murrah Federal Building in Oklahoma City led
to the shearing of main load-bearing columns and resulted in the collapse
of nearly half of the building’s structure, which resulted in a large number
of casualties. In this incident, it was estimated that 87% of the people in
the collapsed portion died (153 out of 175), whereas only 5% of the people
in the uncollapsed portion died (10 out of 186) (Federal Emergency Management Agency 1996), indicating that, had the structure better resisted the
blast loads and not experienced a collapse, the casualties would have been
a lot fewer.
r Even when the structural system survives, blast loads can cause the disintegration of walls and windows, creating secondary fragments that will
generate fatalities and injuries in the adjacent rooms.
In this chapter, methods are described for retrofitting structural components
to enhance their blast resistance and for assessing the protection the retrofits
provide.
Much of this chapter is devoted to describing retrofit techniques to protect
reinforced concrete (RC) columns. This was done because so much can be done
to prevent the catastrophic damage observed for some types of RC columns when
subjected to blast. This information, along with that for the other components, is
presented in a summarized form due to space limitation.
445
446
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
17.2 RETROFIT OF COLUMNS
Typically, columns are the component whose integrity is key to sustaining the
capability of a building’s structural system to survive a blast load. As such, attention to ensuring their survival should have the highest priority when the blast
resistance of a building is to be upgraded.
17.2.1 Reinforced Concrete Columns
Soon after the bombing of the Alfred P. Murrah Building, it was shown that highfidelity physics-based (HFPB) finite element (FE) models would have predicted
the failure of the bearing columns nearest the point of detonation, and the resulting progressive building collapse (Crawford, Wesevich, et al. 1995; Crawford,
Malvar, et al. 1996; Crawford, Bogosian, and Wesevich 1997). These models
also suggested that the use of composite wraps, or steel jackets, could have increased the columns’ blast resistance sufficiently to prevent their failure, and
thus, the building’s collapse.
The concept of jacketing columns had also been suggested by the Federal Emergency Management Agency (Federal Emergency Management Agency
1996), which indicated that retrofit techniques used to resist seismic loads could
be extended to resist blast loads. But it was not until 1999 (Crawford 2001) that,
in a series of three blast tests conducted by the Defense Threat Reduction Agency
(DTRA), it was demonstrated that steel and fiber-reinforced polymer (FRP) jacketing could be used to significantly enhance the blast resistance of RC columns.
Test Data These first three blast tests were conducted on columns located on
the ground floor of a full-scale four-story RC building (Figure 17.1); they are
designated as Tests 1 to 3 in the figure. The same blast load on identical columns
(i.e., except for the retrofit) was used in all three tests, the results being shown in
Figure 17.2.
The Test 1 column’s performance under these conditions is shown in Figure
17.2a, where a diagonal shear failure can be observed—which is not surprising, given the relatively wide tie spacing of this conventionally designed column. However, when this type of column design is retrofitted with a steel jacket
(Test 2), or carbon FRP (CFRP) hoop wraps and vertical CFRP strips (Test 3),
it can survive the same load in nearly pristine condition (Malvar, Morrill, and
Crawford 1999).
Based on these initial results, an extensive series of full-scale column components was tested using various types of FRP retrofit designs. The results from the
blast tests were augmented with a companion program of ten laboratory tests of
column specimens identical to those tested by blast, but loaded statically with hydraulic rams (Figure 17.3). These rams were applied in such a way as to approximate the deformation patterns observed in the blast tests. Aramid FRP (AFRP),
glass FRP (GFRP), and CFRP composites were used successfully in these tests,
although CFRP has shown better durability characteristics (American Concrete
RETROFIT OF COLUMNS
447
Figure 17.1 Location of initial blast tests of unretrofitted and retrofitted RC columns
(tests were conducted using a four-story building constructed using a cast-in-place conventionally designed RC flat-slab framing system; figure shows structure after Test 1
(Photo courtesy of DTRA)
Institute 2002, Malvar 1998a), and its higher stiffness can provide more efficient
confinement with less material (Malvar, Morrill, and Crawford 2004).
Design/Analysis Tools One of the major outcomes of the test program was
the development and validation of a procedure for the design of FRP wrap and
steel jackets to retrofit RC columns to enhance their blast resistance. This procedure can quickly assess the performance of both conventional and retrofitted
RC columns subjected to blast loads, and it provides an easily used means to
design an FRP wrap or steel jacket to retrofit an RC column so that it can survive
a specific blast threat (Malvar, Morrill, and Crawford 1999; Morrill, Malvar, and
Crawford 1999; Morrill, Malvar, and Crawford et al. 2000; Morrill, Malvar, and
Crawford et al. 2004; Perea and Morrill 2002; Kersul and Sunshine 2002). The
retrofit design procedure was embodied in the code CBARD (Crawford, Malvar,
and Ferritto et al. 2003). For composites, the design procedure determines the
number of FRP hoop wraps needed (and possible vertical FRP reinforcement)
for a given blast load. For steel jackets, it determines the required jacket thickness, given the proposed jacket geometry.
448
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
(a)
(b)
(c)
Figure 17.2 Depiction of results from full-scale blast testing of retrofitted and unretrofitted RC columns (Photo courtesy of DTRA): (a) Unretrofitted column (Test 1),
(b) Steel jacket column (Test 2), (c) FRP wrapped column (Test 3).
RETROFIT OF COLUMNS
449
Figure 17.3 Quasi-static laboratory test demonstrating the highly ductile behavior
achievable for an RC column retrofitted with CFRP wrap to prevent shear failure (Photo
courtesy of DTRA)
450
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.4 Measured versus predicted resistance function for an RC column (as computed by CBARD), which is given in the form of lateral pressure applied versus lateral
deflection at mid-height (the added capacity beyond the third hinge formation is provided
by compression-membrane behavior)
In addition to the blast tests, as mentioned, static laboratory tests (Figure 17.3)
with loading conditions that simulated the deformations observed in the blast
tests were conducted to measure the flexural resistance of the columns (Morrill
et al. 2001). Static testing cannot replicate the part of the behavior related to
the initial highly transient response driven by the short-duration pressure loading of the blast (e.g., the apparent material strengthening at high strain rates),
or the wrapping of the blast loads around the column, but it can yield data that
can be used to verify the blast design procedure. For example, these data provide demonstrative evidence of the importance of compression-membrane (C-M)
forces (Figures 17.4 and 17.5), demonstrate the effectiveness of FRP wrap at enhancing shear resistance (Figure 17.5), and provide load-deflection information
for validating design tools (e.g., CBARD).
Although retrofitting columns with steel jackets or composite wraps is reminiscent of seismic retrofits, there are basic differences. For example, apparent
material strengthening occurs at the high strain rates induced by the blast loading (Malvar 1998b, Malvar and Ross 1998), a phenomenon that, for concrete,
has been linked to moisture content at lower strain rates (below about 1/s) and to
inertial effects above that, and that is typically neglected during seismic loading.
Also, the deformed shape of building columns under blast loads usually includes
(a)
(b)
Figure 17.5 Response of conventional and CFRP hoop retrofitted 14-inch square RC
columns (the axial load in the CFRP column initially increases with increasing lateral
deflection due to compression membrane then decreases to 100 kips when the test was
stopped): (a) Axial load, (b) Lateral load
452
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.6 Classic form for resistance function (represented by the middle line) pertaining to the fully ductile response of an RC column; this resistance only develops if the
diagonal shear capacity always exceeds it
two inflection points and three plastic hinges (Figure 17.3), as compared to a
single inflection and two plastic hinges for seismic loading. This is an important difference, as the deformed shape with two inflexion points allows for the
development of C-M forces, which can significantly increase the lateral resistance of the column, typically anywhere between 20% and 100%, depending on
the column’s geometry and the rigidity of the boundary conditions (Crawford,
Malvar, and Morrill et al. 2001). This C-M behavior develops due to the vertical
(axial) constraint at the top of the column provided by the inertia of the supported structure under high loading rates. Figure 17.6 shows schematically the
flexural resistance after the formation of three hinges and the increase provided
by the C-M resistance developed subsequently. C-M behavior is typically not
present during seismic events; it is a phenomenon associated with both the form
and rapidity of the blast response and the presence of significant mass above the
column.
Procedure for Retrofitting RC Columns The key concept employed in
retrofitting an RC column pertains to preventing shear and other brittle forms of
failure under the blast loading, by ensuring that the shear capacity of the column
exceeds its flexural strength (i.e., including its C−M capacity), and therefore
RETROFIT OF COLUMNS
453
ensuring that the column dissipates the most energy possible via flexure. For example, in Figure 17.6, if the shear capacity is depicted by the top line (demarked
with the solid squares), then the complete flexural behavior will be allowed to
develop. This latter situation ensures that the brittle shear failure response is
avoided altogether (i.e., over the whole range of the column’s response), and instead ensures a ductile flexural response up to the point where P-δ failure occurs.
When the shear capacity of the column is too low (e.g., if in Figure 17.6 the shear
capacity is represented by the lowest line, demarked by the open triangles), this
can be remediated by employing steel jackets or composite wraps to shift this line
upward—for example, to the location of the high shear resistance line. In the case
of composites, hoop wraps in general are sufficient to provide the required shear
capacity and ensure ductile behavior, and vertical sheets or strips are not needed.
Moreover, vertical composite reinforcements tend to rupture at low column lateral displacement, often before compression membrane is developed—that is,
at a deflection of half the column depth, or less; see DSWA (Defense Special
Weapons Agency 1998) or Malvar and Wesevich (1996)—whereas the hoop
wraps can remain intact during large lateral deflections (see Figure 17.3).
Description of Design Methodology. This methodology contains a number
of unique features that distinguishes it from other design procedures, such as
those for retrofitting columns to improve their seismic capacity and traditional
American Concrete Institute (ACI) design. These features include:
r Recognition that some RC columns, even those designed for regions of high
seismicity, are prone to fail in shear and/or experience catastrophic loss of
axial capacity under a blast load.
r Using resistance functions (e.g., as shown in Figure 17.4) specially tailored
to the column’s geometry and the retrofit concepts employed.
r Focusing the retrofit design on achieving a fully ductile response mechanism (e.g., as demonstrated in Figure 17.5) for the lateral response of the
column. A primary goal in designing the retrofit is to have a column that is
fully ductile under lateral load.
r Using a best-estimate response-prediction model for selecting design parameters and estimating capacities of retrofitted and unretrofitted columns.
r Using a single-degree-of-freedom (SDOF) model to predict responses to
blast loads for both unretrofitted and retrofitted RC columns.
Blast loads may produce several types of response indicating various deficiencies, or lack thereof, in the column’s behavior. The deficiencies likely to be
encountered include:
r Lack of Diagonal Shear Capacity: The column can fail in shear with little
plastic behavior, hence dissipating little energy. This is a worst-case condition for a column subjected to a blast load, and will likely be the predominant reason for retrofitting columns. This deficiency is sometimes referred
454
r
r
r
r
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
to, although not preferably, as diagonal tension failure rather than diagonal
shear failure.
Lack of Flexural Ductility: Good practice would dictate that columns should
achieve a full measure of flexural ductility for blast-resistant design—a full
measure being reaching their deflection limit, which is related to keeping
the column’s axial demand/capacity ratio less than 1 (one). Deflections associated with this level of ductility (i.e., full measure) are typically on the
order of a large fraction of the column’s width. This may not be practical for
wide, short columns, in which case, one must be content with prevention of
shear and other brittle failure modes.
Lack of Direct Shear Capacity: This condition is uncommon for building
columns. It occurs for columns with high diagonal shear capacity and high
flexural resistance. This condition is of particular concern for columns having low ratios of height to depth, which is common in parking structures
underneath multistory buildings, and for columns with steel jackets. It is
often not practical or easy to enhance the direct shear capacity, but significant energy can be dissipated by the direct shear response itself so that an
acceptable behavior may be achieved without resorting to any extraordinary
retrofit measure.
Lack of Moment Capacity: If the resistance function is low or has little area
under it (i.e., its integral is small, indicating little energy absorption in the
response), little energy will dissipate, and unacceptable lateral deflections
may ensue. This may be mitigated by adding vertical composite strips (FRP
retrofits), a steel jacket, or steel strips (steel retrofits). In these cases, close
attention should be paid to ensuring that the resulting enhanced diagonal
and direct shear demands are met.
Lack of Confinement: This condition is of most concern for large columns
having high axial loads that are subjected to a large blast load. In such a
situation, the lateral response to the blast is likely to be quite small, but the
shock load may be large enough to cause extensive cracking in the concrete.
In this case, if the column has insufficient ties to maintain confinement of
the concrete and prevent rebars buckling, then the axial resistance of the
column may be severely compromised.
The design methodology presented in CBARD encompasses these situations,
and provides a means for designing either FRP wrap or steel jackets to address
all of these deficiencies.
17.2.2 Steel Columns
The retrofit of steel columns is relatively simple, as compared to the design process employed for RC columns. Generally, the main concerns for steel columns
are that they are likely to lack sufficient shear capacity near their supports, and
that they may not have enough flexure resistance to preclude excessive lateral
RETROFIT OF COLUMNS
455
deformation. Retrofit concepts for blast can often employ concepts similar to
those for enhancing the static load resistance of a steel column.
There are considerable blast effects data for steel columns and subassemblies comprising a column fixed at its base with a set of girders framing into
it at a floor level above, and the connections used (Crawford, Magallanes, and
Morrill 2005). These data have provided insight into the behaviors that may
be cause for concern in a blast environment and that may require mitigation.
The data have also been used to validate HFPB finite element models and simplified engineering models for the analysis and design of steel framing systems for blast loads (Morrill, Crawford, and Magallanes 2007, Magallanes and
Morrill 2008).
Damage Modes. In retrofitting steel columns to enhance their resistance to
blast loads, three basic modes of response should be considered:
r Mode 1: Excessive lateral deformations
r Mode 2: Joint failures, for example, for baseplate connections and column
splices that are near the point of detonation
r Mode 3: Fracture or excessive localized deformation of web and flange
The first two modes are relatively straightforward in terms of using retrofit
concepts to mitigate the risk of their occurrence, which are primarily related
to the lateral loading engendered by the blast. The Mode 3 failure, however,
is peculiar to blast where the rapid loading produced by the blast produces a
shear failure along the K-line, or causes the web and flanges to deform excessively. Results from a blast test illustrating some aspects of these failure
modes, taken from one of a series of tests of wide flange segments, are shown in
Figure 17.7.
Retrofit designs for addressing these potential failure mechanisms are likely
to require careful consideration, and no presupposed methodology or approach
can be considered adequate to ensure that all the salient issues at a particular site
are addressed adequately.
Retrofitting. Mode 1 failures can be addressed in relatively conventional
ways by strengthening of the member (e.g., by the addition of cover plates).
Both local and global deformation of the section may need to be addressed, depending on the blast scenarios considered. Concerns for excessive deformation
for light sections may be particularly difficult to address. The retrofit approach
would follow much the same strategy as would be used to retrofit a girder to
take a higher load. This might involve the use of cover plates, stiffeners, knee
braces, and gusset plates. The main concern in developing the retrofit design
will be ensuring that ductile behavior is maintained throughout the response,
especially with regard to the added shear demand caused by blast type loadings at the column’s supports. Alternatively, a blast shield may be installed in
front of the column to reduce its response and negate or lessen the need for
retrofit.
(a)
(b)
(c)
Figure 17.7 Examples of failure modes for steel columns: (a) Section view, illustrating
Mode 1 P-δ failure, (b) Connection view, illustrating concerns with fracture (Mode 3
failure), (c) Baseplate view, depicting a Mode 2 failure (Photo courtesy of DTRA)
RETROFIT OF COLUMNS
457
Addressing Mode 2 failures is a little less straightforward, but they can be
approached in several ways. These failures are related to a general lack of capability in the various sorts of connections used in conjunction with deploying
steel columns. Of these, the failure of column base plates and splices is of most
concern. The retrofits for addressing Mode 2 failures are likely to take a variety
of forms. The general approach employed in reducing the vulnerability to Mode
2 failures is to ensure that the connection does not represent a weak spot. For
example, using a cross tie to remove reliance on the base plate’s capacity. The
other main concern is to ensure that plastic hinges are able to form in the column
as a whole, so as to achieve a fully ductile lateral response for the column.
Mode 3 failures are most easily addressed by taking advantage of the short
duration of the blast load and the benefits that can be derived from inertia in this
situation. A retrofit concept embodying this approach is depicted in Figure 17.8.
Here the interstices between the section’s flanges are filled with concrete, which
mitigates problems associated with localized bending of both the flanges and
webs (Crawford and Lan 2005). This may be augmented with additional webs,
as shown in the figure, if additional shear capacity is desired.
Another means to protect the column from web breach (i.e., in the case of a
wide flange section) is to place a concrete or other similarly substantive panel
in front of the web so that the airblast cannot impinge directly on the web. This
could be in the form of a conventional cladding (Crawford, Magallanes, and
Morrill 2005).
Figure 17.8 Example of the infill concept for preventing web and flange failure
(Mode 3) for steel sections
458
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
17.3 RETROFIT OF WALLS
Bearing walls, like columns, are a key structural component supporting the gravity loads of a building system, and the loss of bearing walls can also lead to overall collapse. Unlike columns, however, bearing walls can sustain major damage
without compromising their function of carrying the gravity loads. These walls,
in essence, offer a continuous form of load path redundancy, which allows them
to be locally breached by a nearby charge and still perform their primary function of holding the structure up. In contrast to columns, walls can also provide
a last line of defense for the building’s occupants from the effects of blast, and
the failure of both bearing and nonbearing walls can result in deadly secondary
fragmentation likely to produce significant casualties as their debris are blown
through a building.
Walls are found in a wide variety of strengths, combine a wide range of materials, vary substantially in quality of construction, and often present considerable
difficulty in estimating their robustness as a structural system. In the following
discussion, the behaviors of walls subjected to out-of-plane static and blast loads
are examined. These are different from seismic applications, where walls are
generally used more to provide shear resistance, and therefore subjected primarily to in-plane loads.
Composites, polymers, and conventional materials have been used to enhance
the blast resistance of masonry walls (Malvar, Crawford, and Morrill 2007).
FRPs have been used to enhance the resistance of RC and stud walls. Sheetmetal panels and blast shields are useful in enhancing the blast resistance of any
wall. Examples of the use of some of these retrofit concepts are described below.
17.3.1 Masonry Walls
The retrofitting of masonry walls with and without reinforcement to enhance
their blast resistance has been widely studied. Those composed of concrete masonry units (CMU) and fired clay bricks will be the focus of the following discussion.
The response of conventional masonry walls to blast loads has been observed
in many full and subscale tests. An example of one of these tests is shown in Figure 17.9. In this figure, two side-by-side identical ungrouted CMU walls were
tested with a moderate-size blast load. The wall to the right was retrofitted (Figure 17.9b); the one to the left was not. Figure 17.10 shows a snapshot depicting
the motions of the unretrofitted wall immediately after a detonation. The performance shown is typical of a masonry wall’s response for a moderate blast load,
in that the CMU blocks remain relatively intact, while the failure of the wall is
largely due to the opening of the masonry’s horizontal joints.
Three approaches to retrofitting walls are described. The first two technologies involve bonding reinforced and/or unreinforced layers of polymer to the
back (and possibly front) faces of a masonry wall to provide both added strength
and ductility for blast load response. The third approach uses some of the same
(a)
(b)
Figure 17.9 Example showing the benefit of using FRP to retrofit a masonry wall to
enhance its blast resistance: (a) Control (left) and retrofitted wall (right) prior to blast
test, (b) Control (left) and retrofitted wall (right) after the test (Photo courtesy of The
Energetic Materials and Research Testing Center)
460
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.10 Example typical of the response expected from a CMU wall struck by a
blast; at the time shown, the CMU is moving at several tens of feet per second (Photo
courtesy of The Energetic Materials and Research Testing Center)
materials as the first two, but in this case they are not bonded to the wall. In this
situation, the strengthening elements, such as FRP, are anchored to the floor and
act by preventing entry of wall debris rather than by wall strengthening.
Fiber-Reinforced Polymer The use of FRP to enhance the strength and blast
resistance of masonry walls has been extensively studied. This would include
FRPs such as CFRP, AFRP, and GFRP.
CMU Studies. Muszynski (1998) and Muszynski and Purcell (2003) bonded
CFRP and AFRP to masonry walls composed of concrete masonry units. Baylot, Dennis, and Woodson (2000), Baylot et al. (2005), and Dennis, Baylot, and
Woodson (2002) conducted experiments on walls using quarter-scale models
built with typical 8-inch-wide CMU. In these tests, three types of retrofitting
schemes were evaluated for ungrouted and for partially grouted walls. The first
retrofit method consisted of 0.04-inch (1 mm) thick GFRP fabric bonded to
the back face of the wall. The second retrofit technique consisted of a twopart sprayed-on polyurea product applied to the back face of the wall. The third
RETROFIT OF WALLS
461
retrofit method used 20-gauge galvanized steel sheet metal attached to the back
of the wall. All the walls failed during the blast testing by separating from the
frame, but the GFRP and the polyurea retrofitted walls were successful in preventing loose fragments within the structure, and were, therefore, considered as
successful retrofits.
Myers, Belarbi, and El-Domiaty (2003, 2004) tested eight ungrouted, unreinforced masonry (URM) walls retrofitted with FRP (epoxy) laminates and
near-surface-mounted (NSM) rods. Two series of walls (l00 and 200 mm or 4
and 8 inches thick) with different slenderness ratios and different strengthening
schemes were tested at varying blast charge weights and standoff distances. The
walls were only supported at the top and bottom, and retrofitted walls failed in
shear at those locations. The FRP laminates were able to contain the wall debris.
Carney and Myers (2003, 2005) tested twelve ungrouted URM walls using a
pressure bag to apply a uniform loading (Phase I), then field-tested two walls under blast (Phase II). Results of the last four Phase I tests, together with the Phase
II tests, are shown in the second paper (Carney and Myers 2005). The walls
were reinforced with GFRP sheets or NSM rods, with and without anchorage.
The anchorages consisted of wrapping the sheets around NSM rods embedded
into the wall’s top and bottom support for the first case, and of extending the
NSM rod three inches into the wall support for the second. Strengthening the anchorages almost doubled the capacity of the walls. In contrast, unanchored NSM
rods did not show any improvement. For FRP reinforced walls, no shear failure
was observed during the static tests, while it was apparent in previous blast tests
(Myers, Belarbi, and El-Domiaty 2003), indicating a potential limitation in the
use of static tests to simulate blast tests.
Stanley, Metzger, and Martinez et al. (2005) used a retrofit system composed
of multiple layers of carbon and glass FRP in an epoxy resin (applied with paint
rollers), placed on the interior and exterior surfaces of the CMU wall test specimen. This system was successful in containing debris but seemed too labor intensive to apply. Stanley, Metzger, and Morrill et al. (2005a) used two layers of a
glass FRP in an epoxy resin applied to the interior surface of the CMU wall test
specimen. This wall experienced complete failure, indicating that stiff retrofit
systems fail catastrophically once their capacity is exceeded.
Simmons (2005) studied the effectiveness of three types of retrofits (E-glass,
22-gauge steel, and an elastomeric polymer) on CMU walls to increase their
resistance to primary fragments from a 120-mm mortar round and a 122-mm
rocket; he found that the E-glass significantly reduced the number of fragments
that passed through the wall.
Brick Studies. Conley and Mlakar (1999) used CFRP (epoxy) strips to reinforce small brick walls loaded with a gas-loaded piston and subjected to one-way
bending. Increases in load capacity of 400% were recorded.
Patoary and Tan (2003) completed static tests using air bags on four built-in
solid masonry brick walls retrofitted with three types of FRP systems: glass, carbon, and low-strength fiberglass woven roving. The walls, 120-mm thick, were
able to sustain deflections up to 20 mm. Specimens with GFRP or CFRP failed
462
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
by edge debonding, while specimens with roving failed by fiber rupture. This
work was followed by out-of-plane tests on 30 masonry walls reinforced with
three different GFRP and CFRP systems (Tan and Patoary 2004).
Retrofitting with Unreinforced or Fiber-Reinforced Polyurea-like Polymers
In some situations, the performance of walls reinforced with FRP composites
is limited by their peak strain capacity, so materials with higher failure strains
might be expected to perform better. Because both fibers and epoxies used for
FRP applications have rather low strains to failure, several researchers started
looking at polyureas, which are materials with purported failure strains of 100%
or more.
These materials have appeared in two forms: one as a spray-on coating that
is applied and bonded directly to the wall’s surface, the other as a sheet polymer
that is placed behind the wall and anchored to the floors. The first system acts in
concert with the masonry by adding tensile capacity. The second system acts as
a catcher system to contain the wall’s debris and is not bonded to the wall. This
latter system is described in Section 17.3.1.3.
Unreinforced. Knox et al. (2000) sprayed the back of concrete block walls
with various unreinforced polymers, in particular polyurethane and polyurea.
These polymers showed exceptional ductility and were able to prevent injurious
wall fragmentation. As indicated earlier, Baylot, Dennis, and Woodson (2000)
and Baylot et al. (2005) and Dennis, Baylot, and Woodson (2002) conducted
experiments on CMU walls using quarter-scale models of typical 8−inch-wide
CMU. Among other retrofits, they used a two-part sprayed-on polyurea product
applied to the back face of the wall, which was successful in preventing loose
fragments within the structure.
Hammons, Knox, and Porter (2002) expanded their application of polyurea
to CMU walls and found it very successful in preventing deadly wall fragmentation. Deflections of up to 610 mm (24 inches) were recorded (the nominal CMU wall width was 8 inches, or 200 mm). Other tests, using a high
pressure spray–applied two-component polyurea (Stanley Martinez, and Koenig
2005b), and a two-component spray-applied polyurea liner (Stanley, Metzger,
and Morrill et al. 2005b) applied to both interior and exterior CMU wall faces,
confirmed such findings, although some tearing of the polymer was observed. In
these applications with large deflections, the complete resistance function may be
mobilized, including the lower resistance beyond compression membrane (due
to the crushing of the concrete compression zone), and the subsequent strengthening due to tensile membrane, enabling the dissipation of significant amounts of
energy.
Davidson, Fisher et al. (2005) studied 12 polymer-reinforced masonry walls
during seven explosive tests. The walls were restrained at the top and bottom
only and responded in a one-way fashion. During the tests, arching due to end
constraints was suspected to have contributed to the loss of the compression face
near the ends. A simple spray overlap of 6 inches (152 mm) was sufficient to
transfer the loads to the frame. Compared to stiff FRP fabric reinforcing, the
RETROFIT OF WALLS
463
spray materials were cheaper, bonded better to the wall, and were easier to anchor
to the frame (see also Davidson, Porter et al. 2004).
Reinforced. These polyurea polymers can also be further strengthened with
continuous fibers, which can provide some additional energy dissipation up to
fiber rupture. Tests were conducted by Crawford and Morrill (2002) to define
the properties of reinforced and nonreinforced polyurea spray-on coatings.
These materials were used to retrofit masonry walls subjected to blast loads.
This is detailed in several studies (Crawford and Morrill 2003a, 2003b; Johnson,
Slawson, Cummins et al. 2003, 2004; Johnson and Slawson 2004; Morrill,
Nicolaisen, and Hutchinson 2005; Piepenburg et al. 2002, 2003; Stanley,
Metzger, and Morrill et al. 2005c; Stanley, Metzger, and Koenig. 2005a).
Piepenburg et al. (2002) used an aramid grid fabric to reinforce a polyurea
spray retrofit of a CMU wall, and were successful in containing the CMU debris. Piepenburg et al. (2003) also used an aramid within two thin sprayed-on
polyurea layers on a CMU wall. Crawford and Morrill (2003a) describe a retrofit
consisting of an aramid laminate that is secured to the wall with a polyurethane
adhesive, and anchored to the diaphragms with energy-dissipating compliant anchorages. Johnson, Slawson, and Cummins et al. (2003, 2004) used an openweave aramid with different fiber lay-ups with a spray-on polyurea. The capacity
of the walls retrofitted with reinforced polyurea was 1.4–2.0 times higher than
that of the walls retrofitted with just polyurea, and 5.5–7.5 higher than that of the
unretrofitted walls. Morrill, Nicolaisen, and Hutchinson (2005) used a polyurea
coating, with and without an aramid fiber mesh, on the back side of brick wall
samples tested quasi-statically in a laboratory environment. These tests were useful in determining complete quasi-static resistance functions, including tensile
membrane.
Stanley, Metzger, and Morrill et al. (2005c) used a spray-applied two-part
polyurea with an inner aramid weaved fabric on a CMU wall, and successfully
stopped the CMU debris. Peak deflections were about 230 mm (9 inches). Stanley, Martinez, and Koenig (2005a) used a non-woven polypropylene geotextile
fabric in between two layers of sprayed-on polyurethane applied to the interior
and exterior surfaces of the CMU wall test specimen. In this case, the retrofitted
wall did fail, although it would appear that the retrofit had provided some amount
of early-time confinement and lateral restraint of the CMU blocks.
Retrofitting with Catcher Systems In contrast to the retrofit schemes described
for walls above, which are related to strengthening them directly, catcher
systems are placed behind the wall to catch any debris that might be generated
by its breakup from a blast. Such a device may be composed of thin sheets
R
of fabric (e.g., Kevlar laminate), sheet metal, or polymer (e.g., as shown in
Figure 17.11).
This form of catcher system may be augmented with cables for cases involving particularly high loads (e.g., impulses of thousands of psi-ms) or where
anchorage conditions prohibit the use of the sheets by themselves. These systems
are designated as cable catcher systems.
464
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Figure 17.11 Example of a thin polymer catcher system being installed (Photo courtesy
of Sherwin-Williams Company)
The intent of these concepts is just to capture the debris of the wall, not to
strengthen the wall. These systems may be constructed in a variety of ways, and
some designs have very large capacities. In addition, these systems are likely
to impart less lateral load to the structural system of the building, which is
significant where collapse of the framing system due to the lateral loading of
the blast is of concern. The debris-catching solution is primarily intended for
RETROFIT OF WALLS
465
Figure 17.12 Example of the type of composite panels that may be used for retrofitting
masonry and stud walls to enhance their blast resistance
non-load-bearing walls whose function is not critical to preventing failure of the
structural system. However, these systems may also be applied to bearing walls,
where the breadth of the wall failure incurred can be spanned by the remaining
structural system.
In 2004, a series of reinforced and unreinforced polymer panels for retrofitting
masonry walls was developed (Crawford, Lan, and Wu 2009). These panels are
delivered to the site in the form of sheets, which are placed behind the wall and
anchored to the floors to hold them in place, as shown in Figure 17.11. These
panels have shown their worth against both vehicle and rocket blast threats in
tests in Israel and the U.S.
Retrofitting with Composite Panels In 2000, large-capacity catcher systems
were introduced and validated in blast tests (Crawford 2001) for retrofitting masonry or stud walls. These systems employed composite panels, like the one
shown in Figure 17.12, that were composed of FRP or steel sheets bonded to a
crushable core material, such as the rigid polyurethane foam shown in the figure.
The panel is placed behind the wall and anchored to the floor, generally using a
compliant anchorage. These systems have been proved successful in blast tests
for both masonry and metal stud walls, and were shown to work for very high
loads if the anchorage is designed appropriately (Crawford 2001, Crawford, Xin,
and Bogosian 2002).
Retrofitting with Conventional Methods Conventional methods (e.g., shotcrete
and dowels) may also be used to retrofit a masonry wall to enhance its blast
466
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
resistance, or just to prevent its debris from propagating into the building. These
systems should provide continuous support to the existing masonry wall to prevent localized breaching. Such systems would be expected to be less efficient
than those already mentioned. Moreover, the generally large quantities of materials used to install these types of retrofits may add to the risks from blast by
introducing considerably more material that could become a source of debris.
17.3.2 Stud Walls
Knox et al. (2000), and Crawford and Morrill (2003a, 2003b) also studied the
retrofit of wood and steel stud walls, using composite materials, which provided
an excellent means for adding blast protection to an existing stud frame building.
In the Crawford and Morrill studies, both fabric and sheet metal composites, such
as the one shown in Figure 17.12, were deployed.
17.4 FLOORS
Considerations of the blast protection afforded by the floors of a building are
likely to involve three kinds of situations, which are related to the location of the
detonation:
r Case 1: Considerations for detonations outside the building, where the detonation is some distance away. Here, enhancing the resistance of the floor
may be unnecessary or may be accomplished using conventional strengthening techniques, mainly because the floor’s response is neither very pronounced nor complex. The actual distance in question depends on the type
of exterior wall, the floor’s properties, and other factors.
r Case 2: Considerations for exterior detonations near the building, with no
wall present (e.g., near a loading dock). Here, an upward shock could be
imparted to a floor above the detonation and downward motion imparted to
the floor below the detonation.
r Case 3: Interior detonation. Here, considerations for two situations are of
issue: the blast above, and the blast below a floor. For an interior detonation,
gas pressures may be of major concern. In this case, adding venting might
be part of the retrofit strategy.
Retrofitting floors to provide protection from blast is likely to be more problematic than retrofitting any other structural component, mainly because of the
large spans and loaded areas, their load resistance being primarily related to gravity (i.e., with limited resistance to uplift), and the possibility that the detonation
may be quite close (e.g., a bomb resting on the floor). Moreover, in contrast to
most other retrofit strategies, the weight of the floor retrofit can be an issue because of the floor’s extent.
FLOORS
467
The following are three retrofit strategies for addressing the three threat situations mentioned above.
r Case 1. When the bomb is outside, the walls are likely to knock the blast
down sufficiently so that the floor response may be ignored—especially
since it’s likely in such a case that the walls would be retrofitted first, and
even less load would reach the floor. Moreover, the main concern is the net
load acting on the floor, which may be negligible, even if the walls fail.
However, in situations where the facades above and below a floor have very
different blast resistance (e.g., ground floors that have no exterior walls), a
substantial net upward (or downward) load on a floor can be generated, and
the situation may require some form of mitigation.
r Case 2. If the bomb is relatively close to the outside of a building and opposite an opening in the wall (or possibly a glass facade), the floor, especially
above the bomb, which would be loaded in its weak direction, may be a
cause of concern. Here, however, the effects of gas pressure can probably
be ignored.
r Case 3. For an interior detonation, both shock and gas pressure components
of the blast may cause concern.
The risks associated with a floor hit by a blast depend greatly on the size and
location of the bomb. For example, is the detonation from a bomb on the floor,
or above/below the floor? This has mostly to do with proximity and whether the
response is mainly localized or involves the whole of the floor, the magnitude of
velocity (shock) imparted to the floor and possible debris generated from it, and
whether the floor is loaded on its strong or weak axis.
Protection for floors against nearby bombs (i.e., within a few feet) is likely to
be impractical. For example, for a bomb resting on a conventional floor, there are
few (or no) reasonably priced ways to prevent the floor’s breach. Where the blast
is above the floor, then attaching a catcher system to the joists to prevent debris
propagation could be a good strategy.
For situations involving a vented blast occurring beneath a floor that has
a reasonable standoff, retrofitting the floor with an FRP may suffice. However, even if the floor remains relatively undamaged, there is likely to be significant shock imparted to it, which can cause severe injuries to those standing on it and launch furnishings, providing another source for injuries. Better
protection can be provided by a blast shield, possibly hung below the actual
floor, to either mitigate or prevent the airblast from striking the floor. In this
circumstance, not only is the floor not damaged, but the velocity imparted
by the shock, resulting in the launch of furnishings and people, is avoided
altogether.
The selection of a retrofit strategy for floors is likely to primarily involve
limiting the risks, rather than eliminating them. Three basic retrofit strategies,
depending on the bomb’s size and location, seem most applicable to addressing
468
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
the vulnerability presented by floors:
r Strengthening the existing floor through the addition of ductility and
capacity.
r Debris containment. This strategy is particularly useful for those floors in
direct contact with or very near to the bomb.
r Shielding. Shielding a floor from the blast load avoids the issue altogether.
This could also include constructing a second floor directly above/below the
existing one that is directly exposed to the threat. Alternatively, the existing
floor may be retrofitted so that it can provide the protection for a new floor.
Retrofit strategies for floors are likely to be much more difficult to implement
than for walls, which would imply that the best strategy here is to prevent a bomb
or an airblast from getting close enough to present a serious risk to the floors,
and to use the walls of the building to protect the floors. Even for a loading dock,
using roll-up doors or some other form of closure can probably deflect the blast
enough so that a floor above it is sufficiently protected.
17.5 BEAMS/GIRDERS/CONNECTIONS
The retrofit strategies pertaining to enhancing the blast protection afforded by
structural components that act as beams, girders, and the connections associated
with them—hereafter collectively referred to as beams—are generally of less
concern than the components already mentioned, because in many ways these
components generally present less of a risk.
Retrofit strategies for beams to accommodate blast loads are likely to be relatively simple and probably similar to those that might be used in conventional
retrofit strategies, especially for blast loads that are generated by charges that
predominantly excite the fundamental modes of the beam.
In addition, whether the blast is pushing up or down on the beam can make
a substantial difference in its response. Many beams are not as well attached in
the upward direction, and may have similar weaknesses in terms of their ability
to resist lateral loads. For closer charges, retrofit strategies are likely to be sitespecific and not easily addressed with simplified tools and conventional retrofit
strategies.
Blast effects data for beams (Crawford, Magallanes, and Morrill 2005,
Magallanes and Morrill 2008), although sparse, indicate that the beams themselves are likely to be less of an issue than their support connections. For external
blast, it is likely that the net loads pushing up/down on the beam will be negligible. However, loading on a beam can be quite localized (e.g., a bomb placed on
the floor immediately above the beam), or broad-based; for example, if applied to
a robust floor system that is strongly coupled to the beams framing it, the beam’s
loading would derive from a large tributary area.
SUMMARY
469
A major concern for beams is their attachment (i.e., connections) to their support (e.g., columns). For many steel beams, especially in nonseismic designs,
this may consist of only a shear tab, or for precast beams only a small ledger to
which it is only lightly attached (e.g., the shear tab shown in Figure 17.7b).
Several retrofit concepts have been developed for seismic applications to improve the anchorage of beams/girders at their supports that would also work in
blast environments, although they may need to be more robust. Another retrofit
strategy that has been used (Crawford and Lan 2005) relies on cabling attached to
the beams to carry the load, should their connection to the columns be damaged.
17.6 STRUCTURAL SYSTEM
The preceding sections addressed ways to retrofit structural components to enhance their survivability and protective capability in the event of a blast. It is
also important, perhaps even more so, to evaluate the structural system’s performance, especially once the retrofits are in place.
However, evaluating the risks to the system and describing retrofit strategies
for addressing them is beyond the scope of this chapter. With regard to assessing risks to the system, however, it is important to recognize that retrofitting
structural components can produce deleterious effects in terms of the system’s
performance. For instance, there would be a need to ensure that the impact of
the additional loading imparted to the system by any enhancement in the protection be evaluated—for example, the added lateral load imparted to the system
by the addition of blast-resistant walls could exacerbate the lateral response of a
building system sufficiently to cause its failure from excessive story drift.
17.7 SUMMARY
Information was presented in this chapter pertaining to retrofitting a facility’s
structural framing system and components to enhance the protection afforded
by them in the event that a blast occurs. Each of the major components of the
structural system was addressed, including columns, walls, beams, and floors.
Several strategies were described for mitigating risks from blast for each type of
component.
17.7.1 Inexact Science
The ideas presented in this chapter concerning the retrofit of structural components to enhance the blast protection afforded by them can only, due to space
limitations, reflect a general notion of strategies and methods that may be pursued in this endeavor. Moreover, only a few of the ways this might be accomplished could be cited, and all the concerns and complexities that might arise in
an actual application could not be addressed. What is intended, however, is to
apprise those desirous of developing retrofit strategies and designs of some of
470
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
the main issues and methods that have been found useful and effective, and to
describe extant test data and studies that are available.
Developing retrofit designs for any structure is often more art than engineering because each situation is likely to be unique, to have issues of unintended
consequences, to raise difficulties of identifying the extant design, and so forth.
In addition, blending with the existing structure, architecture, and functionality
is likely to present a host of conflicting constraints, making the retrofit even more
difficult.
It is likely that any retrofit design will not be able to offer the same certainty
of function as a new design. Moreover, achieving too high a certainty is likely
to result in a poorer level of protection or perhaps none at all—that is, there is
a certain paralyzing effect that can easily be reached by demanding too much
certainty. On the other hand, there is a particular need in this type of engineering
to be certain that the design is not fatally flawed. Adopting a good enough standard may offer the best course for this type of work, the main issue here being
the difficulty of defining such a standard ahead of time (e.g., for the purposes of
generating a contract and design budget).
17.7.2 Complexities
Retrofitting a structure to enhance its resistance to blast may require an inordinate
level of skill and experience. The limited amount of information presented in this
chapter can only allude to some of the considerations that are needed to develop
these retrofits and their integration with the existing structure, and should be
viewed as just an introduction to the subject.
It is important to recognize that, in many cases, the intensity and rapidity of
the blast load generate behaviors substantially different than generally encountered in structural design problems. Moreover, the forces and material damage
incurred are often difficult to determine without using HFPB finite element models, and for most retrofit designs, they are likely to be much higher than ordinarily
dealt with, requiring a familiarity with material and dynamic response behaviors
in this high loading rate regime.
The level of complexity and difficulty encountered in developing retrofit designs is likely to be highly exacerbated when the blast load is from a large,
close charge. These situations can be addressed, but they require considerations
in addition to those that can be alluded to here. Moreover, in these situations,
many simplified engineering models commonly used for blast effects analysis
are likely to be deficient and misleading in their prediction of response.
REFERENCES
American Concrete Institute, Committee 440. 2002. Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures (ACI 440.2R-02).
Farmington Hills, MI: American Concrete Institute.
SUMMARY
471
Baylot J. T., B. Bullock, T. R. Slawson, and S. C. Woodson. 2005. Blast response
of lightly attached concrete masonry unit walls. Journal of Structural Engineering
131 (8): 1186–1193.
Baylot J. T., S. T. Dennis, and S. C. Woodson. 2000. Response of 1/4 scale concrete masonry unit (CMU) walls to blast. In Emerging Technologies in Fluids, Structures, and
Fluid/Structure Interactions, Vol. 1. Proceedings of the 2000 ASME Pressure Vessels
and Piping Conference, Seattle, WA, July 23–27, 2000. New York: American Society
of Mechanical Engineers.
Carney, P. and J. J. Myers. 2003. Shear and flexural strengthening of masonry infill walls
with FRP for extreme out-of-plane loading. Proceedings, Architectural Engineering
Institute Annual Meeting, Austin, TX, 2003.
. 2005. Out-of-plane static and blast resistance of unreinforced masonry
wall connections strengthened with FRP. In 7th International Symposium on FiberReinforced Polymer Reinforcement for Concrete Structures, ACI Symposium Publication 230, 229–248. Farmington Hills, MI: American Concrete Institute.
Conley C. H. and P. F. Mlakar. 1999. Blast resistance of CFRP reinforced masonry. Presented at the 70th Shock and Vibration Symposium, Albuquerque, NM, November
15–19, 1999.
Crawford J. E. 2001. Addressing workplace protection issues more effectively. Proceedings of the Society of American Military Engineers Symposium on Comprehensive
Force Protection, Charlestown, SC, November 1–2, 2001.
Crawford J. E., D. D. Bogosian, and J. W. Wesevich. 1997. Evaluation of the effects of
explosive loads on masonry walls and an assessment of retrofit techniques for increasing their strength. Presented at the 8th International Symposium on the Interaction of
the Effects of Munitions with Structures, Defense Special Weapons Agency, McLean,
VA, April 1997.
Crawford J. E. and S. Lan. 2005. Design and implementation of protection technologies for improving blast resistance of buildings. Presented at the Enhancing Building
Security Seminar, Singapore, March 2005.
Crawford J. E., S. Lan, and Y. Wu. 2009. Design methods for enhancing the blast resistance of masonry walls and windows with bonded and unbonded polymer sheets.
Proceedings of the 8th International Conference on Shock and Impact Loads on Structures, Adelaide, Australia, December 2–4, 2009.
Crawford J. E., J. J. Magallanes, and K. B. Morrill. 2005. Vulnerability of steel framing
systems. Proceedings of the 12th International Symposium on the Effects of Munitions
with Structures, New Orleans, LA, September 26–30, 2005.
Crawford J. E., L. J. Malvar, B. W. Dunn, and D. J. Gee. 1996. Retrofit of reinforced
concrete columns using composite wraps to resist blast effects. Proceedings of the
27th Department of Defense Explosive Safety Seminar, Las Vegas, NV, August 19–22,
1966.
Crawford J. E., L. J. Malvar, J. M. Ferritto, K. B. Morrill, B. W. Dunn, and P. Bong.
2003. Description of Design Software for Retrofitting Reinforced Concrete Columns
to Improve Their Resistance to Blast. (Report TR-01-16.2). Burbank, CA: Karagozian
and Case. Limited Distribution.
Crawford J. E., L. J. Malvar, K. B. Morrill, and J. M. Ferritto. 2001. Composite retrofits to
increase the blast resistance of reinforced concrete buildings. Proceedings of the 10th
472
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
International Symposium on Interaction of the Effects of Munitions with Structures,
San Diego, CA, May 2001.
Crawford, J. E. and K. B. Morrill. 2002. New highly ductile designs for preventing injurious wall debris due to a terrorist bombing. Proceedings of the Protection of Structures
Against Hazards, Singapore, November 2002.
. 2003a. Retrofit techniques using polymers and fiber reinforced polymers for
preventing injurious wall debris. Proceedings of the 6th International Symposium on
Fibre-Reinforced Polymer Reinforcement for Concrete Structures, FRPRCS-6, Singapore, July 8–10, 2003, pp. 1199–1208.
. 2003b. Retrofit techniques for strengthening stud walls subjected to nearby
blast loads. Proceedings, 74th Shock and Vibration Symposium, San Diego, CA, October 27–31, 2003. Limited Distribution.
Crawford J. E., J. Wesevich, J. Valancius, and A. Reynolds. 1995. Evaluation of jacketed columns as a means to improve the resistance of conventional structures to blast
effects. Proceedings, 66th Shock and Vibration Symposium, Biloxi, MS, October
30–November 2, 1995.
Crawford J. E., X. Xin, and D. D. Bogosian. 2002. Retrofit techniques for conventional
walls to prevent their causing injurious debris due to a terrorist bombing. Proceedings
of the 17th International Symposium on Military Aspects of Blast and Shock, Las
Vegas, NV, June 10–14, 2002.
Davidson J. S., J. W. Fisher, M. I. Hammons, J. R. Porter, and R. J. Dinan. 2005. Failure
mechanisms of polymer reinforced concrete masonry walls subjected to blast. Journal
of Structural Engineering 131 (8): 1194–1205.
Davidson J. S., J. R. Porter, R. J. Dinan, M. I. Hammons, and J. D. Connell. 2004. Explosive testing of polymer retrofit masonry walls. Journal of Performance of Constructed
Facilities 18 (2): 100–106.
Defense Special Weapons Agency (DSWA). 1998. Design and Analysis of Hardened
Structures to Conventional Weapons Effects (TM 5-855-1, AFPAM 32-1147(1), NAVFAC P-1080, DAHSCWEMAN-97). Washington, DC: The Departments of the Army,
Air Force, Navy, and The Defense Special Weapons Agency.
Dennis S. T., J. T. Baylot, and S. C. Woodson. 2002. Response of 1/4 scale concrete masonry unit (CMU) walls to blast. Journal of Engineering Mechanics 128 (2):
134–142.
Federal Emergency Management Agency, Building Performance Analysis Team. 1996.
The Oklahoma City Bombing: Improving Building Performance through MultiHazard Mitigation (FEMA 277). Washington, DC: Federal Emergency Management
Agency.
Hammons M. I., K. J Knox, and J. R. Porter. 2002. Polymer materials for blast retrofit
of masonry construction. Proceedings, 30th Department of Defense Explosives Safety
Seminar, Atlanta, GA, August 13–15, 2002.
Johnson C. F. and T. R. Slawson. 2004. Concrete masonry units walls retrofitted
with fiber reinforced elastomeric systems for blast loads. Presented at the 75th
Shock and Vibration Symposium, Virginia Beach, VA, October 27–31, 2004. Limited
Distribution.
Johnson C. F., T. R. Slawson, T. K. Cummins, M. B. Hueste, and L. Beason. 2003.
Concrete masonry units (CMU) static and dynamic wall experiments with elastomeric
SUMMARY
473
retrofits. Presented at the 74th Shock and Vibration Symposium, San Diego, CA, October 27–31, 2003. Limited Distribution.
Johnson C. F., T. R. Slawson, T. K. Cummins, and J. L. Davis. 2004. Concrete masonry
unit walls retrofitted with elastomeric systems for blast loads. Presented at the 24th
Army Science Conference, Orlando, FL, November 29–December 2, 2004.
Kersul A. M. and D. Sunshine. 2002. Structural collapse blast mitigation. Presented at
the 17th International Symposium on Military Aspects of Blast and Shock, Las Vegas,
NV, June 10–14, 2002.
Knox K. J., M. I. Hammonds, J. R. Porter, and T. T. Lewis. 2000. Polymer materials for
structural retrofit. Proceedings, 29th Department of Defense Explosives Safety Seminar, New Orleans, LA, July 18–20, 2000.
Magallanes J. M. and K. B. Morrill. 2008. Fast running model for the residual capacity
of steel columns damaged by blast & fragment loads. Proceedings of the 79th Shock
and Vibration Symposium, Orlando, FL, October 26–30, 2008.
Malvar L. J. 1998a. Durability of composites in concrete. Presented at First International Conference on Durability of Composites for Construction, Sherbrooke, Quebec,
Canada, August 5–7, 1988, pp. 361–372.
. 1998b. Review of static and dynamic properties of steel reinforcing bars. ACI
Materials Journal 95 (5): 609–616.
Malvar L. J., J. E. Crawford, and K. B. Morrill. 2007. Use of composites to resist blast.
Journal of Composites for Construction 11 (6): 601–610.
Malvar L. J., K. B. Morrill, and J. E. Crawford. 1999. CTS-1 Retrofit Designs: For Reducing the Vulnerability of an Office Building to Airblast (Report TR-98-39.2). Burbank,
CA: Karagozian & Case.
. 2004. Numerical modeling of concrete confined by fiber reinforced composites. Journal of Composites for Construction 8 (4): 315–322.
Malvar L. J. and C. A. Ross. 1998. Review of static and dynamic properties of concrete
in tension. ACI Materials Journal 95 (6): 735–739.
Malvar L. J. and J. W. Wesevich. 1996. Peak Compression Membrane Response Enhancements for One-Way and Two-Way Slab Resistance Functions (TM-96-1.2). Burbank,
CA: Karagozian & Case.
Morrill, K. B., J. E. Crawford, and J. M. Magallanes. 2007. “Development of Simplified Tools to Predict the Blast Response of Steel Beam-Column Connections,” Proceedings of the 2007 ASCE/SEI Structures Congress, Long Beach, CA, May 16–19,
2007.
Morrill K. B., L. J. Malvar, and J. E. Crawford. 1999. Retrofit design procedure for existing RC buildings to increase their resistance to terrorist bombs. Proceedings of the 9th
International Symposium on Interaction of the Effects of Munitions with Structures
(IEMS), Berlin, Germany. May 3–7, 1999.
Morrill K. B., L. J. Malvar, J. E. Crawford, and S. W. Attaway. 2000. RC column and
slab retro-fits to survive blast loads. Proceedings of the ASCE Structures Congress,
Philadelphia, PA, May 8–10, 2000.
Morrill K. B., L. J. Malvar, J. E. Crawford, and J. M. Ferritto. 2004. Blast resistant design
and retrofit of reinforced concrete columns and walls. Presented at ASCE Structures
Congress, Nashville, TN, May 23–25, 2004.
474
RETROFIT OF STRUCTURAL COMPONENTS AND SYSTEMS
Morrill K. B., L. J. Malvar, J. E. Crawford, G. Hegemier, and F. Seible. 2001. Fullscale testing of reinforced concrete column retrofits to resist blast loads. Proceedings
of the 10th International Symposium on Interaction of the Effects of Munitions with
Structures, San Diego, CA, May 2001.
R
Morrill K. B., K. Nicolaisen, and T. Hutchinson. 2005. Laboratory Testing of Kevlar
©
Reinforced Polyurea (Report TR-04-52.1). Burbank, CA: Karagozian & Case.
Muszynski L. C. 1998. Explosive field tests to evaluate composite reinforcement of concrete and masonry walls. In Fiber Composites in Infrastructure, Proceedings of the
Second International Conference on Composites in Infrastructure, Tucson, Arizona,
January 5–7, 1998, pp. 277–284.
Muszynski L. C. and M. R. Purcell. 2003. Use of composite reinforcement to strengthen
concrete and air-entrained concrete masonry walls against air blast. Journal of Composites for Construction 7 (2): 98–108.
Myers J. J., A. Belarbi, and K. A. El-Domiaty. 2003. Blast resistance of un-reinforced
masonry walls retrofitted with fiber reinforced polymers. Proceedings for The North
American Masonry Society 2003 Annual Meeting, Clemson, South Carolina, pp.
312–323.
. 2004. Blast resistance of FRP retrofitted un-reinforced masonry (URM) walls
with and without arching action. The Masonry Society Journal 22 (1): 9–26.
Patoary M. K. H. and K. H. Tan. 2003. Blast resistance of prototype in-built masonry
walls strengthened with FRP systems. Proceedings of the 6th International Symposium
on Fibre-Reinforced Polymer Reinforcement for Concrete Structures, FRPRCS-6, Singapore, July 8–10, pp. 1189–1198.
Perea, A. and K. B. Morrill. 2002. Component high explosive testing of bare and
retrofitted columns. Presented at the 17th International Symposium on Military Aspects of Blast and Shock, Las Vegas, NV, June 10–14, 2002.
.Piepenburg, D., R. Martinez, R. Del Frate, and K. Morrill. 2002. Blast Resistance Upgrade for Glazing, Window Frame, and Wall Systems—Non-Load Bearing In-Fill
Walls with Blast Retrofits applied to the Wall, Window Frame and Glaze (Report No.
SAIC-03/1001). San Diego, CA: Science Applications International Corporation.
. 2003. Blast Resistance Upgrade for Glazing, Window Frame, and Wall Systems – Load Bearing Walls with Blast Retrofits applied to the Wall, Window Frame and
Glaze (Report No. SAIC-03/1003). San Diego, CA: Science Applications International
Corporation.
Simmons, L. 2005. Arena fragment experiments using mortars. Presented at the 76th
Shock and Vibration Symposium, Destin, FL, October 30–December 4, 2005.
Stanley, M., R. Martinez, and J. Koenig. 2005a. UL-Like Testing of Commercial Off-theShelf Products that Enhance the Blast and Ballistic Resistance of Structures, Quick
Look Report 6 (New Mexico Tech, Energetic Materials Research and Testing Center,
TR-05-08). Burbank, CA: Karagozian & Case. Limited Distribution.
. 2005b. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 7 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-05-14). Burbank, CA:
Karagozian & Case. Limited Distribution.
Stanley, M., J. Metzger, R. Martinez, and J. Koenig. 2005. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance the Blast and Ballistic Resistance of
SUMMARY
475
Structures, Quick Look Report 5 (New Mexico Tech, Energetic Materials Research and
Testing Center, TR-05-03). Burbank, CA: Karagozian & Case. Limited Distribution.
Stanley, M., J. Metzger, K. B. Morrill, and R. Martinez. 2005a. UL-Like Testing of
Commercial Off-the-Shelf Products that Enhance the Blast and Ballistic Resistance of
Structures, Quick Look Report 2 (New Mexico Tech, Energetic Materials Research and
Testing Center, TR-04-48). Burbank, CA: Karagozian & Case. Limited Distribution.
. 2005b. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 3 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-04-50). Burbank, CA:
Karagozian & Case. Limited Distribution.
. 2005c. UL-Like Testing of Commercial Off-the-Shelf Products that Enhance
the Blast and Ballistic Resistance of Structures, Quick Look Report 4 (New Mexico
Tech, Energetic Materials Research and Testing Center, TR-04-55). Burbank, CA:
Karagozian & Case. Limited Distribution.
Tan K. H. and M. K. H. Patoary. 2004. Strengthening of masonry walls against outof-plane loads using fiber-reinforced polymer reinforcement. Journal of Composites
for Construction 8 (1): 79–87.
Handbook for Blast-Resistant Design of Buildings
Edited by Donald 0. Dusenberry
Copyright 0 2010 by John Wiley & Sons, Inc. All rights reserved.
INDEX
A
acetone peroxide, 26
Alfred P. Murrah Federal Building, 17,
22, 27, 173, 446, 447
alternative load path, 8
Amman, 25
analysis methods, 243, 244, 256, 389
Baker-Strehlow-Tang (BST), 179
blast curves, 179
component modeling, 249
computational fluid dynamics (CFD),
178
connections, 252
effects of preload, 251
envelope design approaches, 265–67
equivalent static loads, 112
explicit dynamic finite elements,
256
finite elements, 120, 245, 269
for moment frames, 246
for steel-braced frames, 247
multi-degree-of-freedom, 254, 256,
268
P-I diagrams, 253, 267
precast tilt-up construction, 247
precision of, 15
ray-tracing procedures, 209
reinforced masonry, 247
resistance factors, 112
semi-empirical, 210
shear walls, 247
simplified, 4, 251
single-degree-of-freedom, 9, 95, 123,
245, 254, 268
sophisticated approaches, 9
TNO Multi-energy Method (MEM),
179
antiterrorism, 102
appurtenances, 287
assessment
asset value, 18, 31
asset value ratings, 34
design-base threat, 51
interior explosion risk, 298
naturally occurring explosions, 58
risk, 4, 28, 38, 61, 242, 243
risk assessment models, 20
site-specific risk, 18
threat, 18, 51, 61
threat mapping, 36
threat ranking scale, 31
threat rating, 31
vulnerability, 18, 34, 54
vulnerability checklist, 37
vulnerability rating, 35, 37
Atlanta Centennial Olympic Park, 25
B
Baghdad, 23
base isolation, 292
of mechanical systems, 337
Beirut U.S. Embassy, 22
Beirut U.S. Marine headquarters, 22
blast consultants
qualifications, 14
blast design purposes, 89
blast design strategies, 88
blast environment, 5
blast loads, 4, 186
air-induced ground shock, 354
air-shock-induced motions, 356
angle of incidence, 189, 192
calculated by computational fluid
dynamics (CFD), 183
combining shocks from multiple
transmission media, 357
confidentiality, 13, 30
confined explosions, 199
477
478
INDEX
blast loads (Continued )
design base threat, 29, 52, 61, 71,
243
designations, 29
determination of, 55
direct-induced ground shock, 355
drag coefficients, 197
dynamic wind pressure, 185
empirical method of calculation,
183
factors of safety, 64
in reentrant corners, 293
in-structure shock, 353
leakage, 7, 207, 350, 351
methods of prediction, 175
on overhangs, 293
ray-tracing procedures, 209
rear wall, 198
reflected, 187
roof, 197
side wall and roof, 193
sources for guidance, 62
verification, 269
blast parameters
prediction of, 172
surface bursts, 174
blast pressures
as a function of standoff, 310
clearing, 162
deflagration, 162, 163
detonation, 163
propagation through ducts, 352
rarefaction, 162
reflected, 162
blast response, compared to seismic
response, 12
blast simulators, 154, 155
blast walls, 12, 328
blast wave. See shock wave
blowout panels, 282
bombs, delivery mechanisms, 62
building codes, 79
building envelope, 46
balanced or capacity design, 267
connections, 272
convex vs. concave, 294
curtain walls, 276
design objective, 264
operable windows, 278
point and cable supported curtain
walls, 278
robustness, 53
building programming,
dispersal/distribution of, 54
business continuity, 47
C
catch systems, 7, 287, 441, 464
chemical, biological, and radiological
(CBR) control, 48
collateral damage, 45
components
blast walls, 12
frangible, 6
non-structural, 9
sacrificial elements, 10
shielding, 10
concrete
beam detailing, 376
beam reinforcement, 377
beam splice locations, 378
beam-column joint detailing, 378
column detailing, 373
column reinforcement, 374
column retrofit, 453
column splice locations, 376
composite, 291, 388
confinement, 455
detailing, 371
detailing for direct shear, 375, 378
detailing for shear strength, 379
direct shear, 455
direct shear failure, 143
dynamic design stresses, 139
dynamic increase factor, 137
effect of strain rate on compressive
strength, 131, 140
effect of strain rate on tensile strength,
132
exposed columns, 288
failure modes, 141, 369
flat slabs, 243, 290
flexure, 455
mechanical properties for design,
138
plain, 134
punching shear failure, 143
reinforcement, 133, 136, 378
INDEX
reinforcing steel splices, 372
scabbing, 144, 249
shear, 454
slab connection detailing, 380
slab detailing, 379
slab reinforcement, 380, 389
spalling, 144, 249
strength reduction factors, 145
stress-strain relationships, 129, 130
stress-strain relationships in tension,
131
wall connection detailing, 381
wall detailing, 381
wall reinforcement, 381
concrete reinforcement, stress-strain
relationships, 132
confidentiality, 13, 30, 70
consequence evaluation, 44
consequence management, 18, 42, 57
consequence mitigation, 45
constitutive models, reinforced concrete,
132
consultants
qualifications, 14
teaming, 18, 31, 45, 49, 52, 56, 59, 68,
73, 266
control centers, security, 76
counterterrorism, 27
courtyards, 278
Crime Prevention through Environmental
Design (CPTED), 76
critical assets, 33
critical functions
dispersal, 18
D
damage expectations, based on levels of
protection, 90
damage level for petrochemical facilities,
106
Dar es Salaam, 22
defense in depth, 32, 75
deflagration. See blast pressures
deflection limits, 87
concrete components, 108
concrete members, 93
doors, 109
ductility ratio, 91, 94, 95, 96, 101, 103,
107, 108, 109, 113, 114, 393
479
high or heavy damage, 114
limitations on, 114
masonry components, 108
medium damage, 113
steel components, 107
steel members, 93
support rotation, 91, 93, 94, 95, 96,
101, 103, 107, 108, 109, 113, 114,
393
design approaches, 18, 424
balanced or capacity design, 271
design basis threat, 18, 29, 58
design philosophy
design to budget, 73
design to threat, 71
detonation. See blast pressures
disproportionate collapse. See
progressive collapse
doors, 107, 279, 281
ductility, 244
ductility ratio, 93, 126, 386, 388, 390,
396, 397, 407, See also deflection
limits
dynamic design stresses, 126
dynamic increase factor, 112, 124
concrete, 137
steel, 125
E
East Beirut U.S. Embassy annex, 22
egress, 12, 265
stairwell enclosures, 299
electrical system, 68
emergency response, 12, 44, 46, 48
evacuation, rescue, and recovery systems,
30, 48, 79, 265
emergency elevator system, 348
emergency lighting, 48
evaluation of, 54
fire, 49
fire alarm, detection, exiting, and
annunciation, 48
fire suppression, 48
redundancy, 49, 55
survivability, 49
video surveillance, 48
explosions
accidental, 58, 59
below grade effects, 290
480
INDEX
explosions (Continued )
boiling liquid expanding vapor
explosion (BLEVE), 170
bursting pressure vessels, 178
causes, 57
confined, 165
deflagration, 184
intentional, 60
interior, 174, 175–77, 298
leakage, 264
pressure vessel burst, 169
scaled ranges, 249
surface bursts, 174
vapor cloud, 168–69, 172, 178–80
venting of internal explosions, 175
explosive events, 19, 21, 248
management, 54
explosives
acetone peroxide, 26
ammonium nitrate and fuel oil
(ANFO), 164
C-4, 26, 59
high, 165, 172
low, 165
nitromethane, 165
Propellants and pyrotechnics, 165
RDX, 26, 59
Semtex, 26, 59
TNT, 26, 59, 163
triacetone triperoxide, 26, 27
F
facade. See building envelope
facility function redundancy, 55, 56
facility siting, 51
facility staffing and operations, 341
failure modes, 127
breach and spall, 433
diagonal tension, 370
direct shear, 249, 370, 433
flexure, 128, 141, 369, 429
flexure and axial load, 128, 144
fracture, 252
instability, 128, 371
lateral-torsional buckling, 392
local buckling, 392
membrane action, 370
P-Delta effects, 251
punching shear, 243
scabbing, 144
shear, 128, 142, 432
spalling, 144
steel columns, 457
fenestration, 273
fiber-reinforced polymer, 442
fire, 81, 163, 302, 305, 345
control center, 348
detection and alarm systems, 349
smoke-control systems, 349
flammable agents, 81
fragmentation, 248, 264, 329
armor-piercing fragment, 235
catch systems, 287
constrained fragments, 224
damage to roofs, 237
damage to various materials, 238
debris missiles, 7
design fragment parameters, 227
from pressure vessel bursts, 170
Gurney energy, 219
impact damage, 229
K constant, 224
K1 and K2 constants, 226
K3 penetration factor, 236
loadings, 216
material toughness, 226
non-armor-piercing fragment, 236
of glass, 46
penetration, 230
penetration into concrete targets,
234
perforation, 236
primary, 216, 217
range prediction, 229
secondary, 217, 219
spalling, 237
steel targets, 232
structural missiles, 10, 50
THOR constants, 231
THOR equation, 230
THOR equation variables for various
target materials, 233
trajectory, 228
unconstrained fragments, 220
velocity, 219, 227
frangible components, 6
Friedlander equations, 248
functional redundancy, 18, 48
INDEX
G
Glasgow International Airport, 24
glass, 274
analysis software, 275
H
Hopkinson scaled pressures and
impulses, 153
HVAC systems, 46
hydrocode, 175
I
improvised explosive devices (IED), 22
person-borne, 24
personnel, 61
vehicle-borne, 22, 30, 61
information sensitivity, 13
J
Jakarta JW Marriott Hotel, 23
K
Karachi Marriott Hotel, 24
Karachi U.S. Consulate, 23
kinetic energy, 10
L
large deformations, 9
levels of protection, 71, 90
load combinations, 392
load factors, 9
London, 27
louvers, 282
M
Madrid, 24, 25, 27
masonry
breach and spall, 433
damage expectations, 430
detailing, 435
effect of coursing, 426
failure modes, 427
grout, 426
mortar, 426
performance assumptions, 423
reinforcement, 425, 436
reinforcing steel splice capacity,
426
retrofit recommendations, 441
481
splice locations, 426
strength, 425
support connections, 439
unreinforced, 424, 440
wall analysis, 439
wall detailing, 438
mass, effect on blast response, 10
materials, constitutive relationships,
247
mechanical systems, 68
anchorage, 338, 359
design philosophy, 334
electrical systems, 345
emergency elevator system, 348
emergency power systems, 347
fire protection systems, 345
general announcing communication
systems, 349
HVAC, 46
HVAC and plumbing systems,
342
in hardened spaces, 342
level of protection, 335
lighting systems, 347
location of, 335, 341
plenums, 299
protection of, 265
segregation, 76
seismic vs. blast, 333, 336
shock induced by the structure,
336
shock isolation, 337, 339
system fragility, 337
testing, 333
mitigation strategies, 19, 60
momentum, 10
N
Nairobi, 22
O
occupant injuries, 8
operational security, 65
P
Paddy’s Bar, 23
peer review, 69, 157
Pentagon, 17
penthouses, 291
482
INDEX
performance criteria, 88, 386, See also
deflection limits
damage level, 97, 98
development bases, 99
explosive safety, 99, 101
facades, 270
for antiterrorism, 102, 103, 105
for petrochemical facilities, 105
publications, 100
performance goals, 87, 89
performance verification, 149
by analysis, 156
by testing, 150
perimeter protection, 310
active wedge, 321, 323
anti-ram systems, 33, 53, 317
beam barriers, 321, 323
berms and ditches, 325, 327, 328
bollards, 320, 322
cable-based systems, 324
costs, 312
crash modeling, 318
crash testing, 317
Crime Prevention through
Environmental Design, 33
pedestrian control barriers, 326
planters and surface barriers, 325,
328
personnel screening, 75
petrochemical facilities, 105
physical security, 65
planning, secure facility, 20
plenums, 298
plumbing system, 68
polyurea-like polymers
reinforced, 464
unreinforced, 463
polyvinyl butyral interlayer, 280, 299
post-explosion conditions, 77
potential energy, 11
pressure-impulse diagram, 95, 96
progressive collapse, 11, 46, 53, 67, 77,
257, 285, 303, 384, 385
alternative load path, 12
European guidance, 259
U.S. guidance, 259
protection category, descriptions,
102
pyrotechnic reaction, 168
Q
quality assurance, 69
R
recovery and contingency planning,
56
redundancy
of assets, 75
of building systems, 34
reinforcing steel, 136
effect of strain rate on tensile strength,
134, 140
effect of strain rate on yield strength,
133, 140, 424
splice capacity, 426
strain-rate effects, 136
stress-strain relationships, 133
resistance-deflection curves, 92
resource allocation, 71, 73
response criteria. See deflection limits
response limits, 87
response parameters, 91
restoration of operations, 47
retrofit
analysis and design tools, 448
beams and connections, 469
column test data, 447
concrete columns, 447
fiber reinforced polymer, 449, 450,
452, 460, 461
floors, 467
masonry walls, 459
polyurea-like polymers, 463
steel columns, 455
steel jacket column, 449, 451
steel sections, 458
stud walls, 467
tests, 451, 461
walls, 459
with composite panels, 466
risk acceptance, 70, 74
risk assessment. See also assessment
risk assessment process, 19
risk management, 19, 40
strategies, 43
risk mitigation, 39
risk rating, 38
risk reduction, 41, 72
Riyadh, 23, 27
INDEX
roofs, 290
concrete, 290
gardens, 291
steel, 290
S
sacrificial elements, 10
safe havens, 300
access to egress, 305
design requirements, 302
location, 304
mechanical systems, 342
screening
of cargo and mail, 76
of personnel, 75
of vehicles, 76
security design guidelines, 82
security zone, 75
Shanksville, Pennsylvania, 17
shielding, 10
shock wave, 163, 170
comparison of free field and reflected
blast loads, 171
from detonation, 164
high explosives, 184
Hopkinson-Cranz scaling method, 173
overpressure, 171
parameters, 170
propagation, 162
scaling, 171
site-specific risk assessment, 18
siting of facilities, 36, 293, 314, 326
balancing hardening with standoff, 310
skylights, 279
stairwell enclosures, 299
standoff, 51, 53, 55, 64, 308
pressures as a function of, 310
standoff military ordnance
rocket-propelled grenade, 30
shoulder fired missiles, 30
steel
column retrofit, 455
composite, 291, 388
connection design, 419
connection ductility, 387
constitutive models, 120
deck, 290
design examples, 398
detailing, 394
483
detailing for failure modes, 394
ductility, 386
dynamic design stress, 391
dynamic increase factors, 125, 390,
391, 420
effect of strain rate on mild steel, 122
material properties, 389
overstrength factor, 387
strain hardening, 126
strain-rate effects, 124, 387
strength increase factors, 390
strength reduction factors, 145
stress-strain relationships, 119
tensile strength, 121
yield strength, 121
strain rates, 9
strength reduction factors, 8
concrete, 145
steel, 145
stress increase factor. See dynamic
increase factor for various materials
stress-strain relationships
concrete, 129
steel, 119, 120
structural behavior, 8, 256
large deformations, 252
P-Delta effects, 11, 251
seismic vs. blast, 242, 266
structural design
general guidance, 65
impacts on facility programming, 68
structure
braced frames, 243
exposed structural systems, 287, 388
influence on blast load, 6, 244
moment frames, 243, 246
support rotation. See deflection limits
T
Taba Hilton, 24
target attractiveness, 47
tensile strength
effect of strain rate on, 121
effect of temperature on, 121
tension membrane, 93
testing
building components, 151
mechanical systems, 333
vehicle barriers, 150
484
INDEX
threat
identification and rating, 18, 28, 30,
298
map, 56, 58
methods to respond, 39
mulit-hazard, 301
probability and consequence rating, 30
sequential events, 301
threat reduction, 57, 72
TNT equivalence, 165
of high explosives, 166
of vapor cloud explosions, 169
triacetone triperoxide, 26
trinitrotoluene. See explosives, TNT
U
Unabomber, 25
utility siting, 53
V
value engineering, 69
vapor cloud explosions. See explosions
vaults, 292
vehicle barrier testing, 150
vehicle screening, 76
vehicle speed, 315
verification, 269
vulnerability
assessment planning, 34
reduction of, 63
vulnerability assessment, 18
W
walls
blowout panels, 282
concrete, 283
exterior, 282
masonry, 286
metal panels, 287
post-tensioned panels, 286
precast panels, 284
progressive collapse resistance, 285
retrofit, 459
secondary, 442
shotcrete, 442
stud, 286, 467
tilt-up construction, 286
windows, 109, 273
anchorage, 440
Department of Defense hazard levels,
110
frames and anchorages, 280
Interagency Security Committee
hazard levels, 110, 111
mullions, 279
operable vs. inoperable, 278
supporting structure, 281
World Trade Center, 17, 22, 23
Y
yield stress
effect of temperature on, 121
effects of strain rate on, 121
